TECHNICAL FIELD

The present disclosure relates to transmission and reception devices processing flexible configurable time-frequency resources. In particular, the present disclosure relates to low latency frame structures and adaptive channel estimation techniques.

BACKGROUND

In mobile communications systems, the time between a data packet generated at a transmitter until the packet arrives at a receiver is referred to as the latency or delay. The latency is composed of several subcomponents such as processing delay, propagation delay, and the so called structural delay. Processing delay is a function of the hardware processing power, propagation delay is a function of the distance between the communicating nodes. Those two delay components are not controllable by the system designer. The structural delay however, is a design aspect. It defines the series of operations the information bits undertake from the moment they are generated at the application layer at the transmitter until the application layer at the receiver. The PHY layer latency can be mainly characterized by the Time Transmission Interval (TTI), which is the shortest time interval a message block from the transmitter can be delivered to the receiver. In the LTE system, the TTI is fixed to 1 millisecond.

In future communication applications there will be a need to support different latency requirements.

SUMMARY

It is the object of the invention to provide a concept for improving flexibility of mobile communication with respect to latency requirements, in particular in mobile

communication systems such as LTE.

This object is achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the figures. A basic idea of the invention is to support a flexible PHY layer framework, which can be easily configured to support different latency requirements, while still being backward compatible with legacy systems, such as LTE. A suitable receiver algorithm and a suitable radio resource management according to this novel PHY layer framework are introduced.

The concept is based on flexible TTI lengths which depends on the service

requirement and propagation delay and receiver complexity. The differentiated Latency of Service provided allows different services to use the PHY layer in a more tailored way to their needs rather than having a rigid PHY specification as present in state-of-the art.

The disclosure covers a range of aspects. The topics presented can be subdivided into four main categories: FDD frame structure (described below with respect to Figures 5 to 8), TDD frame structure (described below with respect to Figures 1 1 to 17), adaptive channel estimation (described below with respect to Figures 9 to 10), and latency prioritized resource management (described below with respect to Figures 18 to 19). The common structure is described below with respect to Figures 2 to 4. In the following description, radio signals in time-frequency domain are defined. Such radio signals may have a basic structure according to LTE standardization as shown in Fig. 1 . This figure shows a schematic diagram of a radio signal 100 in a time-frequency representation according to LTE. The radio signal 100 comprises frequency resources in frequency direction 101 , for example 12 subcarriers 103 and time resources in time direction 102, for example 14 symbols 104. In LTE a number of 14 symbols may span 2 slots 105 in time direction 102 and may have a duration of 1 millisecond. The time- frequency resources are partitioned in resource elements 106. The whole resource elements 106 depicted in Fig. 1 form a resource block. A resource block may have a different number of resource elements 106 depending on the definition of a cyclic prefix and other parameters. Fig. 1 is just an example for such a resource block. A first section 108 of the resource block, e.g. defined by the first three symbols R, 2, 3 in time direction including its frequency resources in frequency direction, may form a control section of the resource block in which resource elements carry control data. A second section 109 of the resource block, e.g. defined by the symbols 4, 5, 6, 7, R, 2, 3, 4, 5, 6 ,7 in time direction including its frequency resources in frequency direction, may form a data section of the resource block in which resource elements carry user data. Besides the control data and the user data each resource block carries a specific pilot pattern having specific pilots R, 107 distributed over the resource block at known positions, i.e. positions known to the receiver. The pilots, also referred to as reference signals R are inserted at fixed locations in time/frequency as shown in Fig. 1. The transmission and reception devices described herein may be implemented in wireless communication networks, in particular communication networks based on mobile communication standards such as LTE, in particular LTE-A and/or OFDM. The transmission and reception devices described herein may further be implemented in a base station (NodeB, eNodeB) or a mobile device (or mobile station or User Equipment (UE)). The described devices may include integrated circuits and/or passives and may be manufactured according to various technologies. For example, the circuits may be designed as logic integrated circuits, analog integrated circuits, mixed signal integrated circuits, optical circuits, memory circuits and/or integrated passives. The transmission and reception devices described herein may be configured to transmit and/or receive radio signals. Radio signals may be or may include radio frequency signals radiated by a radio transmitting device (or radio transmitter or sender) with a radio frequency lying in a range of about 3 Hz to 300 GHz. The frequency range may correspond to frequencies of alternating current electrical signals used to produce and detect radio waves.

The transmission and reception devices described herein may be designed in accordance to mobile communication standards such as e.g. the Long Term Evolution (LTE) standard or the advanced version LTE-A thereof. LTE (Long Term Evolution), marketed as 4GLTE and beyond, is a standard for wireless communication of high-speed data for mobile phones and data terminals.

The transmission and reception devices described herein may be applied in OFDM systems. OFDM is a scheme for encoding digital data on multiple carrier frequencies. A large number of closely spaced orthogonal sub-carrier signals may be used to carry data. Due to the orthogonality of the sub-carriers crosstalk between sub-carriers may be suppressed.

The transmission and reception devices described herein may include MAC (media access control) modules for generating or processing transport blocks. The MAC module implements the MAC protocol sublayer that exists in UE and eNodeB. It is part of the LTE air interface control and user planes. The main services and functions of the MAC sublayer include: Mapping between logical channels and transport channels; Multiplexing and demultiplexing of MAC SDUs (service data units) belonging to one or different logical channels into/from transport blocks (TB) delivered to/from the physical layer on transport channels; scheduling information reporting; Error correction through HARQ (hybrid automatic repeat request) protocol; Priority handling between logical channels of one UE; Priority handling between UEs by means of dynamic scheduling; Transport format selection; and Padding. The MAC protocol sublayer is for example specified in 3GPP TS 36.321 - Evolved Universal Terrestrial Radio Access (E-UTRA); Medium Access Control (MAC) protocol specification.

In order to describe the invention in detail, the following terms, abbreviations and notations will be used:

BS: Base Station, eNodeB, eNB

UE: User Equipment, e.g. a mobile device or a machine type communication device

V2X: Vehicle to Infrastructure

V2V: Vehicle to Vehicle

4G: 4 ^th generation according to 3GPP standardization

5G: 5 ^th generation according to 3GPP standardization

LTE: Long Term Evolution

MTC: Machine Type Communication

BLER: Block Error Rate

FDD: Frequency Division Duplex

TDD: Time Division Duplex

TTI: Transmission Time Interval

MCS: Modulation and Coding Scheme or Set

CSI: Channel State Information

UL: Uplink

DL: Downlink

CQI: Channel Quality Information

MAC: Media Access Control

LA: Link Adaptation

TB: Transport Block

(l)DFT: (Inverse) Discrete Fourier Transform

(l)FFT: (Inverse) Fast Fourier Transform

M2M: Machine to Machine LTE-M: Machine to Machine version of LTE

D2D: Device to Device

RF: Radio Frequency

BEP: Bit Error Probability

MSE: Mean Square Error

SNR: Signal to Noise Ratio

Gl: Guard Interval or Guard Period

According to a first aspect, the invention relates to a transmission device for transmitting a radio signal over a radio channel, the transmission device comprising: a media access control (MAC) module configured to generate a transport block; a modulation and coding module configured to encode and modulate the transport block onto time-frequency resources according to a transmission time interval (TTI) window, wherein the time- frequency resources comprises at least one symbol in time direction and a plurality of subcarriers in frequency direction; a controller, configured to adjust the TTI window in symbol units according to a latency budget and a transmission module, configured to transmit the encoded and modulated transport block of a radio signal carrying the time- frequency resources over a radio channel. Such an adjusting of the TTI window provides the advantage of improved flexibility for transmitting a radio signal with respect to latency requirements.

In a first possible implementation form of the transmission device according to the first aspect, the controller is further configured to determine the latency budget based on the TTI duration and a processing power of the receiving device and a propagation delay of the radio channel.

This provides the advantage that different requirements can be used for defining the latency budget. Hence a high flexibility is guaranteed for communications.

In a second possible implementation form of the transmission device according to the first aspect, the controller is configured to determine the propagation delay based on a timing advance parameter.

This provides the advantage that the timing advance parameter, as a parameter received from the base station, is available and can be easily used for computing the propagation delay. In a third possible implementation form of the transmission device according to the first aspect as such or according to any one of the preceding implementation forms of the first aspect, the time-frequency resources is subdivided into a plurality of frequency bands, each frequency band corresponding to a certain latency requirement.

This provides the advantage that by using different frequency bands each one having its own latency requirement, mobile communication can be performed as flexible as desired. In a fourth possible implementation form of the transmission device according to the third implementation form of the first aspect, the plurality of frequency bands comprise a first latency frequency band, a second latency frequency band and a third latency frequency band, wherein the second latency frequency band comprises a higher number of symbols in time direction than the first latency frequency band, and wherein the third latency frequency band comprising a higher number of symbols in time direction than the second latency frequency band.

This provides the advantage that the symbols can be flexibly assigned to different latency frequency bands.

In a fifth possible implementation form of the transmission device according to the first aspect as such or according to any one of the preceding implementation forms of the first aspect, the modulation and coding module is configured to encode and modulate the transport block into one symbol TTI window.

This provides the advantage that a minimum latency, by using one symbol TTI window, can be implemented.

In a sixth possible implementation form of the transmission device according to the first aspect as such or according to any one of the preceding implementation forms of the first aspect, the transmission module is further configured to transmit a signaling message over the radio channel, wherein the signaling message indicates the number of symbols that the time-frequency resources comprises in time direction or the latency budget. This provides the advantage that the used configuration can be easily signaled to the receiver. In a seventh possible implementation form of the transmission device according to the first aspect as such or according to any one of the preceding implementation forms of the first aspect, the transport block comprises information bits and the modulation and coding module is configured to map the pilot symbols according to a predetermined pilot pattern onto the time-frequency resources.

This provides the advantage that the predetermined pilot pattern is known to the receiver and no extra signaling has to be implemented. In an eighth possible implementation form of the transmission device according to the seventh implementation form of the first aspect, the modulation and coding module is configured to map the control bits at predetermined positions around a center subcarrier onto the time-frequency resources. This provides the advantage that the positions for the control bits are known to the receiver, so a fast decoding can be performed at the receiver.

According to a second aspect, the invention relates to a reception device for receiving a radio signal over a radio channel, the reception device comprising: a reception module, configured to receive a radio signal within a transmission time interval (TTI), the radio signal comprising a time-frequency resources which comprises a known pilot pattern; a channel estimation filter, configured to estimate the radio channel based on a selection of pilots from the known pilot pattern; and a controller, configured to select the pilots from the pilot pattern according to a given latency budget.

Such a selection of the pilots from the pilot pattern according to a given latency budget provides the advantage of improved flexibility for transmitting/receiving a radio signal with respect to latency requirements. In a first possible implementation form of the reception device according to the second aspect, the controller is configured to select pilots from current and past symbols of the pilot pattern in the case of a TTI composed of a single symbol.

This provides the advantage that by using current and past symbols for pilot selection, the latency restrictions can be met. In a second possible implementation form of the reception device according to the second aspect as such or according to the first implementation form of the second aspect, the controller is configured to select the pilots from the pilot pattern which minimize a channel estimation error of the channel estimation filter for the given latency budget.

This provides the advantage that by selecting pilots that minimize the channel estimation error, the accuracy of channel estimation and thus data throughput can be improved.

In a third possible implementation form of the reception device according to the second implementation form of the reception device, the channel estimation error comprises a mean squared error depending on an auto-correlation function of a channel response at the pilots, an auto-correlation function of a channel response at all received elements and a cross correlation function of the channel response at the pilots and the channel response at all received elements.

This provides the advantage that such correlation functions can be easily computed.

According to a third aspect, the invention relates to a transmission device for transmitting a radio signal over a radio channel, the transmission device comprising: a media access control (MAC) module configured to generate a transport block; a modulation and coding module configured to encode and modulate the transport block onto a time-frequency resources, wherein the time-frequency resources comprises a first pilot pattern during a first time interval and a second pilot pattern during a second time interval; a controller, configured to control a mapping of the first pilot pattern and the second pilot pattern onto the time-frequency resources; and a transmission module, configured to transmit a radio signal carrying the time-frequency resources over a radio channel.

Such a controlling of the mapping of the first and second pilot patterns onto the time- frequency resources provides the advantage of obtaining a high accuracy of estimating the autocorrelation function.

In a first possible implementation form of the transmission device according to the third aspect, the first pilot pattern comprises scattered pilots; and the second pilot pattern comprises comb pilots continuously distributed in time direction and preamble pilots continuously distributed in frequency direction. In a second possible implementation form of the transmission device according to the third aspect as such or according to the first implementation form of the third aspect, the controller is configured to periodically switch between the mapping of the first pilot pattern and the second pilot pattern.

In a third possible implementation form of the transmission device according to the third aspect as such or according to any of the preceding implementation forms of the third aspect, the controller is configured to adjust a subcarrier spacing of the time-frequency resources based on a given latency budget and a property of the radio channel, in particular a coherence bandwidth and a coherence-time or a Doppler shift. The Doppler

0 423

shift can be used to determine the coherence time, e.q. as ^: . The RMS (root

Doppler shift

mean square) delay can be used to determine the coherence bandwidth, e.g. as

1 1

or .

50x(RMS Delay) 5x (RMS Delay) Pilots are usually used to estimate the channel's second order statistics. A state-of-the-art estimation method is given as follows: To estimate the coherence time or Doppler shift, the auto-correlation function of the channel's response can be calculated for different time shifts. The estimated auto-correlation function is then mapped into reference curves, to obtain an estimate of the coherence time. An example of the reference auto-correlation curves is the zero-th order Bessel's function of first kind.

Similarly, pilots are usually used to estimate the coherence bandwidth or RMS delay by comparing an estimated auto-correlation function to reference auto-correlation function. For example, for an exponential decaying power delay profile, the reference auto- correlation function is known to be— _: — , where

1+ j2uT _rms kf _sc r _rms is the reference RMS delay, k is the subcarrier shift, and f _sc is the subcarrier spacing. In another possible

implementation, the inverse Fourier transform of the estimated channel's frequency response may be computed. This yields an estimate of the Power Delay Profile (PDP). The RMS delay can then be easily computed from the Power Delay Profile. Besides these, other state-of-the-art methods can also be employed to determine the coherence bandwidth.

Similarly, many state-of-the-art methods can be used to estimate the received signal energy, signal power, or signal to noise ratio, e.g. by using pilots. Usually, the more pilots are used, the less the estimation MSE and the more accurate the estimation will be. In a fourth possible implementation form of the transmission device according to the third implementation form of the third aspect, the transmission device comprises: a channel state information (CSI) interface configured to receive CSI of the radio channel, wherein the controller is configured to determine the property of the radio channel based on the received CSI.

In a fifth possible implementation form of the transmission device according to any of the third or fourth implementation forms of the third aspect, the transmission interface is configured to transmit a signaling message over the radio channel, the signaling message indicating the subcarrier spacing adjusted by the controller.

In a sixth possible implementation form of the transmission device according to any of the third to the fifth implementation forms of the third aspect, the subcarrier spacing has a minimum and maximum threshold depending on the property of the radio channel, in particular the coherence bandwidth and the coherence time or the Doppler shift.

In a seventh possible implementation form of the transmission device according to any of the third to the sixth implementation forms of the third aspect, the controller is configured to maintain a fixed cyclic prefix to symbol duration ratio for different subcarrier spacings.

According to a fourth aspect, the invention relates to a reception device for receiving a radio signal over a radio channel, the reception device comprising: a reception module, configured to receive a radio signal, the radio signal comprising a time-frequency resources which comprises a first pilot pattern during a first time interval and a second pilot pattern during a second time interval, wherein the second pilot pattern comprises comb pilots continuously distributed in time direction and preamble pilots continuously distributed in frequency direction; and a channel estimation filter, configured to determine a coherence time and a coherence bandwidth of the radio channel based on the second pilot pattern.

Such a usage of a first pilot pattern during a first time interval and a second pilot pattern during a second time interval provides the advantage of improved flexibility and accuracy for transmitting/receiving a radio signal with respect to latency requirements. In a first possible implementation form of the reception device according to the fourth aspect, the channel estimation filter is configured to determine the coherence time based on a signal-to-noise ratio (SNR) estimation at pilots of the second pilot pattern. In a second possible implementation form of the reception device according to the first implementation form of the fourth aspect, the channel estimation filter is configured to determine the coherence time based on a selection of the pilots of the second pilot pattern which have an SNR above a predetermined threshold. In a third possible implementation form of the reception device according to the fourth aspect as such or according to any of the first or second implementation forms of the fourth aspect, the reception device comprises a channel state information (CSI) interface, configured to transmit CSI indicating the coherence time and the coherence bandwidth of the radio channel, or containing a ratio of the coherence time to the coherence bandwidth in which the number of bits used to indicate the time/frequency selectivity of the channel is reduced.

In a fourth possible implementation form of the reception device according to the fourth aspect as such or according to any of the preceding implementation forms of the fourth aspect, the first pilot pattern comprises scattered pilots, and the channel estimation filter is configured to estimate channel coefficients of the radio channel based on the first pilot pattern.

In a fifth possible implementation form of the reception device according to any of the third or fourth implementation forms of the fourth aspect, the reception module is configured to receive signaling information on a subcarrier spacing of the time-frequency resources over a control channel.

In a sixth possible implementation form of the reception device according to the fifth implementation form of the fourth aspect, the channel estimation filter is configured to estimate the channel coefficients of the radio signal depending on the signaled information on the subcarrier spacing.

According to a fifth aspect, the invention relates to a transmission device for transmitting a radio signal over a radio channel, the transmission device comprising: a media access control (MAC) module configured to generate a transport block; a modulation and coding module configured to encode and modulate the transport block onto a time-frequency resource block having a predetermined duration, wherein the time-frequency resource block comprises a plurality of downlink (DL) control and data channels, a guard interval and an uplink (UL) control channel which are organized in a time-division duplex (TDD) manner; and a transmission module, configured to transmit a radio signal carrying the time-frequency signal resource block over a radio channel.

Such an organization of the time-frequency resources in a plurality of downlink (DL) control and data channels, a guard interval and an uplink (UL) control channel according to TDD provides the advantage of improved flexibility and accuracy for

transmitting/receiving a radio signal with respect to latency requirements.

In a first possible implementation form of the transmission device according to the fifth aspect, the channels are organized in time as DL control, DL data, a guard interval and UL control, wherein a modulation and coding scheme of a first subframe of a portion of the DL control channel is pre-defined and a modulation and coding scheme of the subsequent subframes is variable.

In a second possible implementation form of the transmission device according to the first implementation form of the fifth aspect, the modulation and coding scheme of the subsequent subframes is variable according to channel conditions or according to latency requirements.

In a third possible implementation form of the transmission device according to the second implementation form of the fifth aspect, the UL control channel exists in each frame or exists once for each multiple frames.

In a fourth possible implementation form of the transmission device according to any of the first to the third implementation forms of the fifth aspect, the DL control channel of a current frame comprises information about the modulation and coding schemes and/or a number of symbols used by the control channel of a subsequent frame and/or a modulation and coding scheme of the current or subsequent data channel.

In a fifth possible implementation form of the transmission device according to the fifth aspect as such or according to any of the preceding implementation forms of the fifth aspect, the control channels comprise HARQ and CQI information. In a sixth possible implementation form of the transmission device according to the fifth aspect as such or according to any of the preceding implementation forms of the fifth aspect, the transmission device comprises a controller, configured to adjust a length of the guard interval based on a distance to a base station or on a round-trip propagation delay.

In a seventh possible implementation form of the transmission device according to any of the fourth to the sixth implementation forms of the fifth aspect, the DL data channel contains more than one TTI. According to a sixth aspect, the invention relates to a transmission device for transmitting a radio signal over a radio channel, the transmission device comprising: a media access control (MAC) module configured to generate a transport block; a modulation and coding module configured to encode and modulate the transport block onto a time-frequency resources, wherein the time-frequency resources comprises at least one symbol in time direction and a plurality of subcarriers in frequency direction; a controller, configured to subdivide the time-frequency resources into a plurality of frequency bands and/or time periods, each frequency band and/or time period corresponding to a certain latency; and a transmission module, configured to transmit a radio signal carrying the time-frequency resources over a radio channel.

Such a subdivision of the time-frequency resources into a plurality of frequency bands and/or time periods, where each frequency band and/or time period corresponds to a certain latency provides the advantage of improved flexibility and accuracy for

transmitting/receiving a radio signal with respect to latency requirements.

In a first possible implementation form of the transmission device according to the sixth aspect, the plurality of frequency bands comprise a first latency frequency band, a second latency frequency band and a third latency frequency band, wherein the second latency frequency band comprises a higher number of symbols in time direction than the first latency frequency band, and wherein the third latency frequency band comprises a higher number of symbols in time direction than the second latency frequency band.

In a second possible implementation form of the transmission device according to the sixth aspect as such or according to the first implementation form of the sixth aspect, the controller is configured to assign each frequency band and/or time period to a specific transmission time interval (TTI). In a second possible implementation form of the transmission device according to the sixth aspect as such or according to any of the preceding first implementation forms of the sixth aspect, the controller is configured to support different TTIs within the time-frequency resources simultaneously.

In a third possible implementation form of the transmission device according to the sixth aspect as such or according to any of the preceding first implementation forms of the sixth aspect, the controller is configured to assign at least a portion of each frequency band and/or time period to a specific user. Usually there are several users which use the whole band/period.

In a fourth possible implementation form of the transmission device according to the third implementation form of the sixth aspect, the controller is configured to determine a TTI assigned to a frequency band and/or time period of a user based on latency requirements, propagation delay and processing delay of a service provided to the user.

In a fifth possible implementation form of the transmission device according to the fourth implementation form of the sixth aspect, the transmission module is configured to transmit a signaling message to each user indicating the TTI assigned to the frequency band and/or time period of the user.

In a sixth possible implementation form of the transmission device according to the fifth implementation form of the sixth aspect, the controller is configured to group users of similar TTI in a same user class.

In a seventh possible implementation form of the transmission device according to the sixth implementation form of the sixth aspect, the controller is configured to determine a scheduling priority factor for a user being assigned to a frequency band and/or time period which belongs to a different user class.

In an eighth possible implementation form of the transmission device according to the seventh implementation form of the sixth aspect, the scheduling priority factor is computed as a function of a ratio of the TTI duration of the different user class to the user's TTI duration. In a ninth possible implementation form of the transmission device according to the fifth implementation form of the sixth aspect, the controller is configured to scatter the resources used for the same TTI throughout the time-frequency resources. In a tenth possible implementation form of the transmission device according to the fifth implementation form of the sixth aspect, the controller adaptively modifies the ratio of frequency bands or time periods of different TTI classes according to the traffic load of each class of TTIs. According to a seventh aspect, the invention relates to a multicarrier system, where the number of the multicarrier symbols within a TTI is variable and chosen depending on a latency parameter determined by a latency budget which takes into account structural delay, processing power and propagation delay which is especially characterized by the timing advance parameter; where a message signaling the number of the multicarrier symbols within a TTI and/or the latency parameter is sent from the transmitter to the receiver.

Such a variable number of multicarrier symbols within a TTI that can be chosen depending on a latency parameter provides the advantage of improved flexibility and accuracy for transmitting/receiving a radio signal with respect to latency requirements.

In a first possible implementation form of the multicarrier system according to the seventh aspect, channel estimation at the receiver is based on an adaptive window, i.e. the pilots used for channel estimation are selected depending on the signaled message.

In a second possible implementation form of the multicarrier system according to the seventh aspect as such or according to the first implementation form of the seventh aspect, a single symbol TTI corresponds to the shortest latency supported by the system. In a third possible implementation form of the multicarrier system according to the seventh aspect as such or according to any of the preceding implementation forms of the seventh aspect, the modulation and coding (MCS) scheme is chosen such that the transport block fits into the minimum number of TTIs.

In a fourth possible implementation form of the multicarrier system according to the seventh aspect as such or according to the first implementation form of the seventh aspect, the time/frequency dimensions of the channel estimation window takes into account the channel's coherence time and bandwidth.

In a fifth possible implementation form of the multicarrier system according to any of the first, third or fourth implementation forms of the seventh aspect, the window includes the maximum number of future symbols in the same TTI.

In a sixth possible implementation form of the multicarrier system according to any of the first to the fifth implementation forms of the seventh aspect, the optimal channel estimation filter is causal for the current symbol in each TTI if the TTI is composed of a single symbol.

In a seventh possible implementation form of the multicarrier system according to any of the first to the sixth implementation forms of the seventh aspect, the number of pilots inside a window is fixed and corresponds to the latency budget allocated to the receiver's computation speed.

In an eighth possible implementation form of the multicarrier system according to the seventh aspect as such, initial synchronization is based on dedicated resource elements, and after initial synchronization, the dedicated resource elements are used for transmitting data, and online fine synchronization is based on pilots.

According to an eighth aspect, the invention relates to a multicarrier system with given bandwidth, where the transmitter periodically switches the pilot patterns from being scattered to preamble with comb (continuous time and frequency) pilots; the transmitter adaptively determines the subcarrier spacing according to the latency deadline and the coherence time and frequency estimated using the comb pilots; the determined subcarrier spacing or symbol duration is signaled to the receiver (e.g. via a control channel). Such an adaptive determination of the subcarrier spacing provides the advantage of improved flexibility and accuracy for transmitting/receiving a radio signal with respect to latency requirements.

In a first possible implementation form of the multicarrier system according to the eighth aspect, the subcarrier spacing has a minimum and maximum threshold depending on the channel. In a second possible implementation form of the multicarrier system according to the eighth aspect as such or according to the first implementation form of the eighth aspect, the channel estimation at the receiver is done depending on the signaled subcarrier spacing or symbol duration.

In a third possible implementation form of the multicarrier system according to the eighth aspect as such or according to the first implementation form of the eighth aspect, the receiver estimates the SNR at the pilot subcarriers, selects the subcarriers which have a SNR higher than a certain threshold in order to compute the coherence time.

In a fourth possible implementation form of the multicarrier system according to the eighth aspect as such or according to the first implementation form of the eighth aspect, for OFDM systems, the CP to symbol duration ratio is maintained for different subcarrier spacings, in case of TDD, knowledge of coherence time/frequency at the transmitter is obtained through channel reciprocity, or in case of FDD, the choice of the subcarrier spacing has a minimum and maximum threshold which correspond to block and flat fading assumptions of channel estimation.

According to a ninth aspect, the invention relates to a communication system, where a time-frequency resource block with fixed duration, contains DL control and data channels, a guard interval and UL control channel organized in a TDD manor; the channels are organized in time as DL data, DL control, and UL control.

Such a selection of the time-frequency resource block provides the advantage of improved flexibility and accuracy for transmitting/receiving a radio signal with respect to latency requirements.

In a first possible implementation form of the communication system according to the ninth aspect, the control channel contains HARQ and CQI information.

In a second possible implementation form of the communication system according to the ninth aspect, the base station subdivides the cell into different zones with different distances to the base station. In a fourth possible implementation form of the communication system according to the third implementation form of the ninth aspect, each zone has a frame structure with a different guard interval according to the round trip propagation delay. In a fifth possible implementation form of the communication system according to the ninth aspect as such or according to the first implementation form of the ninth aspect, the control channel of the current frame contains information which corresponds to the subsequent frame.

In a sixth possible implementation form of the communication system according to the ninth aspect, the data channel (PDSCH) can be subdivided into several short-time TTIs according to the latency requirement of the user.

In a seventh possible implementation form of the communication system according to the fifth implementation form of the ninth aspect, the smallest TTI period is 1 symbol. For low latency applications, the UL control channel may exist in each frame or can be only present once each several frames.

According to a tenth aspect, the invention relates to a wireless communication system, wherein the system supports different TTI lengths simultaneously; the base station receives requests from users for services with different latency requirements; the base station computes the suitable TTI window length in time for each user; taking into account the latency requirement of the service, the propagation delay (distance from base station), and the processing delay (user computation power); the base station informs each user with the suitable TTI window length to fulfill its latency requirement through a control message; the base station groups each set of users into classes with similar TTI window length.

Such a computation of the suitable TTI window length provides the advantage of improved flexibility and accuracy for transmitting/receiving a radio signal with respect to latency requirements. In a first possible implementation form of the mobile communication system according to the tenth aspect, after each fixed period of time the base station allocates each class of users into a frequency band/time period which is dedicated to a certain TTI window length.

In a second possible implementation form of the mobile communication system according to the first implementation form of the tenth aspect, the bands/periods may be scattered throughout the time/frequency resource. In a third possible implementation form of the mobile communication system according to the first or second implementation form of the tenth aspect, within each band/period, several TTI windows may exist, however a priority is granted to a certain TTI length (for e.g. prioritized round robin).

In a fourth possible implementation form of the mobile communication system according to the tenth aspect, the band/period for each TTI length is adaptively modified according to the traffic load of each latency class.

BRIEF DESCRIPTION OF THE DRAWINGS

Further embodiments of the invention will be described with respect to the following figures, in which:

Fig. 1 shows a schematic diagram of a radio signal 100 in a time-frequency representation according to LTE;

Fig. 2 shows a block diagram illustrating a transmission device 200 according to an implementation form;

Fig. 3 shows a block diagram illustrating a transmission device 300 according to implementation form;

Fig. 4 shows a block diagram illustrating a reception device 400 according to an implementation form;

Fig. 5 shows a schematic diagram of time-frequency resources 500 partitioned into a first latency frequency band 501 , a second latency frequency band 502 and a third latency frequency band 503 according to an implementation form;

Fig. 6 shows a block diagram illustrating a transmission device 600 according to an implementation form;

Fig. 7 shows a schematic diagram of time-frequency resources 700 illustrating selected pilots RS and non-selected pilots RN according to an implementation form; Fig. 8 shows two exemplary performance diagrams 800a, 800b illustrating mean square error over SNR for different Doppler frequencies;

Fig. 9 shows a schematic diagram of time-frequency resources 900 partitioned for a High Doppler channel or a high RMS delay channel according to an implementation form;

Fig. 10 shows a schematic diagram of time-frequency resources 1000 in which different resource blocks carry different pilot patterns according to an implementation form; Fig. 1 1 shows a schematic diagram of a TDD radio frame 1 100 having downlink (DL), uplink (UL) and guard interval (Gl) sections according to an implementation form;

Fig. 12 shows a schematic diagram of a self-contained TDD radio frame 1200 according to an implementation form;

Fig. 13 shows a schematic diagram of a radio cell 1300 which is subdivided into concentric circles according to an implementation form;

Fig. 14 shows a schematic diagram of a radio signal 1400 including a first frame 1401 followed by a second frame 1402 according to an implementation form;

Fig. 15 shows a schematic diagram of a frame 1500 in which the PDSCH 1501 is subdivided into several TTIs according to an implementation form; Fig. 16 shows a schematic diagram illustrating an exemplary frame structure 1600 for low latency communication according to an implementation form;

Fig. 17 shows a schematic diagram illustrating an exemplary frame structure 1700 for low latency streaming applications according to an implementation form;

Fig. 18 shows a schematic diagram of time-frequency resources 1800 in which time is subdivided into different periods according to a latency aware frequency-segmented resource allocation (LAFRA) approach according to an implementation form; and Fig. 19 shows a schematic diagram of time-frequency resources 1900 in which time is subdivided into different periods according to a latency aware time-segmented resource allocation (LATRA) approach according to an implementation form. DETAILED DESCRIPTION OF EMBODIMENTS In the following detailed description, reference is made to the accompanying drawings, which form a part thereof, and in which is shown by way of illustration specific aspects in which the disclosure may be practiced. It is understood that other aspects may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present disclosure is defined by the appended claims.

It is understood that comments made in connection with a described method may also hold true for a corresponding device or system configured to perform the method and vice versa. For example, if a specific method step is described, a corresponding device may include a unit to perform the described method step, even if such unit is not explicitly described or illustrated in the figures. Further, it is understood that the features of the various exemplary aspects described herein may be combined with each other, unless specifically noted otherwise. Fig. 2 shows a block diagram illustrating a transmission device 200 for transmitting a radio signal over a radio channel according to an implementation form. The transmission device includes a transport block generator 201 , e.g. a media access control (MAC) module as described below, a channel coding unit 202 and a modulation unit 203, e.g. a modulation and coding module as described below. The transport block generator 201 generates a transport block and the channel coding and modulation units 202, 203 encode and modulate the transport block onto time-frequency resources 204, e.g. a resource block 100 as described above with respect to Fig. 1 .

The TTI is composed of a certain number of multicarrier symbols. The duration of each symbol is determined by the subcarrier spacing (plus the Cyclic Prefix (CP) if exists). For example, LTE is composed of 14 symbols. Each symbol has a duration of around 71 μβ, which corresponds to a subcarrier spacing of 15 KHz and a CP 4.7 s. Note that throughout this disclosure, the term TTI is used in a flexible meaning. In LTE, TTI is always fixed to a fixed duration, however in this disclosure TTI is variable. During this TTI, the time/frequency grid is loaded with information which can only be decoded once all the 14 symbols arrive at the receiver. Hence the total structural latency is 14 ^χ 71 μ8=1000μ8. In one implementation, the information bits will be loaded into a TTI which can be formed of multiples of a single symbol. This implies that the granularity of PHY latency becomes a single symbol instead of multiples of 14 symbols.

Additionally, the TTI is not restricted to a single fixed value as in LTE (14 symbols), but can be adaptively modified according to the latency requirement. Hence, a so called adaptive TTI window is introduced, where the number of symbols per TTI can be varied according to the latency requirements.

In an advanced implementation form, the transmitter can modify the subcarrier spacing within a TTI. By changing the subcarrier spacing, the distance between pilots in absolute time/frequency can be changed. Accordingly, the channel estimation error may increase/decrease. For example, in high speed channels such as vehicular

communications, the channel changes quickly in time, hence the subcarrier spacing should be increased so that the pilots get closer to each other in time in order to improve channel estimation.

At the receiver side, the fact that the TTI window is variable implies that the receiver should also be able to decode the information within a TTI in a variable way. In particular, this means that channel estimation, demodulation, channel decoding, CRC check, etc should be all finalized within a TTI.

In one implementation form, the receiver selects the optimal pilots which can be used for estimating the channel according to the channel's second order statistics. The pilot selection window ensures that the selected pilots are the ones which can minimize the Mean Squared Error (MSE). For example, for a high speed wireless links with short multipath echoes, it is more suitable to select pilots in the frequency direction than the time direction, since the channel changes quickly in time, hence the channel response at the old pilots are probably going to be outdated compared to pilots separated in the frequency axis.

In one implementation, a self-contained TDD frame structure is presented. In this variant, the DL control, data and the UL ACK are all contained in a single frame. The cell may be subdivided into different zones each having a different guard interval adapted to the maximum expected round trip propagation delay within the zone. The DL data channel may also be subdivided to several short TTIs according to the latency requirement. The existence of the UL ACK may or may not exist in each frame depending on the channel quality, the DL desired throughput and the level of feedback needed.

The TTI is closely related to the radio resource management since it defines the resource block which defines the smallest unit of allocating resources to a user. Since the smallest time unit of a resource block can be a single symbol, the time resolution of resource allocation becomes finer. Since there are different TTI lengths within a cell, the base station should be able to take the latency of each user into account when allocating a certain resource. The spectrum may be subdivided into different bands where in each band a certain TTI window has a priority over other users having different TTI windows. In one implementation form, the time-frequency resources 204 may implement the resources of a multicarrier system, where the number of the multicarrier symbols within a TTI is variable and chosen depending on a latency parameter determined by a latency budget which takes into account structural delay and propagation delay which is especially characterized by the timing advance parameter. A message signaling the number of the multicarrier symbols within a TTI and/or the latency parameter may be sent from the transmitter to the receiver. Channel estimation at the receiver may be based on an adaptive window, i.e. the pilots used for channel estimation may be selected depending on the signaled message. A single symbol TTI may correspond to the shortest latency supported by the system. The modulation and coding (MCS) scheme may be chosen such that the transport block fits into the minimum number of TTIs. The time/frequency dimensions of the channel estimation window may take into account the channel's coherence time and bandwidth. The window may include the maximum number of future symbols in the same TTI. In case that the TTI is composed of a single symbol the optimal channel estimation filter is causal for the current symbol in each TTI. The number of pilots inside a window is fixed and may correspond to the latency budget allocated to the receiver's computation speed. Initial synchronization may be based on dedicated resource elements, and after initial synchronization, the dedicated resource elements may be used for transmitting data, and online (fine) synchronization may be based on pilots. In one implementation form, the time-frequency resources 204 may implement the resources of a multicarrier system with a given bandwidth, where the transmitter periodically switches the pilot patterns from being scattered to preamble with comb (continuous time and frequency) pilots. The transmitter may adaptively determine the subcarrier spacing according to the latency deadline and the coherence time and bandwidth estimated using the pilots. The determined subcarrier spacing or symbol duration may be signaled to the receiver (e.g. via a control channel). The subcarrier spacing may have a minimum and maximum threshold depending on the channel. The channel estimation at the receiver may be done depending on the signaled subcarrier spacing or symbol duration. The receiver may estimate the SNR at the pilot subcarriers, may select the subcarriers which have a SNR higher than a certain threshold in order to compute the coherence time. For OFDM systems, the CP to symbol duration ratio may be maintained for different subcarrier spacing. In case of TDD, knowledge of coherence time/frequency at the transmitter may be obtained through channel reciprocity, or in case of FDD, the choice of the subcarrier spacing may have a minimum and maximum threshold which correspond to block and flat fading assumptions of channel estimation. In one implementation form, the time-frequency resources 204 may implement the resources of a communication system, where a time-frequency resource block with fixed duration, contains DL control and data channels, a guard interval and UL control channel organized in a TDD manner. The channels may be organized in time as DL data, DL control, and UL control. The control channel may contain HARQ and CQI information. The base station may subdivide the cell into different zones with different distances to the base station. Each zone may have a frame structure with a different guard interval according to the round trip propagation delay. The control channel of the current frame may contain information which corresponds to the subsequent frame. The data channel (PDSCH) can be subdivided into several short-time TTIs according to the latency requirement of the user. The smallest TTI period may be 1 symbol. For low latency applications, the UL control channel may exist in each frame or can be only present once each several frames.

In one implementation form, the time-frequency resources 204 may implement the radio resources of a communication system, where the system supports different TTI lengths simultaneously. The base station may receive requests from users for services with different latency requirements. The base station may computes the suitable TTI window length in time for each user; taking into account the latency requirement of the service, the propagation delay (distance from base station), and the processing delay (user computation power). The base station may inform each user with the suitable TTI window length to fulfill its latency requirement through a control message. The base station may group each set of users into classes with similar TTI window length. After each fixed period of time the base station may allocate each class of users into a frequency band/time period which is dedicated to a certain TTI window length. The bands/periods may be scattered throughout the time/frequency resource. Within each band/period, several TTI windows may exist, however a priority is granted to a certain TTI length (for e.g. prioritized round robin). The band/period for each TTI length may be adaptively modified according to the traffic load of each latency class.

Fig. 3 shows a block diagram illustrating a transmission device 300 according to an implementation form. The transmission device 300 is an implementation of the

transmission device 200 described above with respect to Fig. 2. It includes a media access control (MAC) module 301 , a modulation and coding module 303, a controller 305 and a transmission module 307.

In a first variant of the transmission device 300, the MAC module 301 is configured to generate a transport block 302. The modulation and coding module 303 is configured to encode and modulate the transport block 302 onto time-frequency resources 306 according to a transmission time interval (TTI) window 304. The time-frequency resources 306 comprise at least one symbol in time direction and a plurality of subcarriers in frequency direction. The controller 305 is configured to adjust the TTI window 304 in symbol units according to a latency budget. The transmission module 307 is configured to transmit the encoded and modulated transport block of a radio signal 308 carrying the time-frequency resources 306 over a radio channel.

The controller 305 may determine the latency budget based on the TTI duration and a processing power of a receiving device and a propagation delay of the radio channel. The controller 305 may determine the propagation delay based on a timing advance parameter. The time-frequency resources 306 may be subdivided into a plurality of frequency bands, each frequency band corresponding to a certain latency. The plurality of frequency bands may comprise a first latency frequency band, e.g. a first latency frequency band 501 as shown in Fig. 5, a second latency frequency band, e.g. a second latency frequency band 502 as shown in Fig. 5 and a third latency frequency band, e.g. a third latency frequency band 503 as shown in Fig. 5. The second latency frequency band 502 may comprise a higher number of symbols in time direction than the first latency frequency band 501 . The third latency frequency band 503 may comprise a higher number of symbols in time direction than the second latency frequency band 502.

The modulation and coding module 303 may encode and modulate the transport block 302 into one symbol TTI window. The transmission module 307 may transmit a signaling message over the radio channel, wherein the signaling message indicates the number of symbols that the time-frequency resources comprise in time direction or the latency budget. The transport block 302 may comprise information bits. The modulation and coding module 303 may map the pilot symbols according to a predetermined pilot pattern onto the time-frequency resources 306. The modulation and coding module 303 may map the control bits at predetermined positions around a center subcarrier onto the time- frequency resources 306.

The first variant of the transmission device 300 is further described below with respect to Figures 5 to 8.

In a second variant of the transmission device 300, the media access control (MAC) module 301 is configured to generate a transport block 302. The modulation and coding module 303 is configured to encode and modulate the transport block 302 onto a time- frequency resources 306, wherein the time-frequency resources 306 comprise a first pilot pattern during a first time interval and a second pilot pattern during a second time interval. The controller 305 is configured to control a mapping of the first pilot pattern and the second pilot pattern onto the time-frequency resources 306. The transmission module 307 is configured to transmit a radio signal 308 carrying the time-frequency resources 306 over a radio channel.

The first pilot pattern may comprise scattered pilots. The second pilot pattern may comprise comb pilots continuously distributed in time direction and preamble pilots continuously distributed in frequency direction. The controller 305 may periodically switch between the mapping of the first pilot pattern and the second pilot pattern. The controller 305 may adjust a subcarrier spacing of the time-frequency resources 306 based on a given latency budget and a property of the radio channel, in particular a coherence bandwidth and a coherence-time or a Doppler shift. The transmission device 300 may include a channel state information (CSI) interface configured to receive CSI of the radio channel. The controller 305 may determine the property of the radio channel based on the received CSI. The transmission module 307 may transmit a signaling message over the radio channel, the signaling message indicating the subcarrier spacing adjusted by the controller 305. The subcarrier spacing may have a minimum and maximum threshold depending on the property of the radio channel, in particular the coherence bandwidth and the coherence time or the Doppler shift. The controller 305 may maintain a fixed cyclic prefix to symbol duration ratio for different subcarrier spacing. The second variant of the transmission device 300 is further described below with respect to Figures 9 and 10.

In a third variant of the transmission device 300, the MAC module 301 is configured to generate a transport block 302. The modulation and coding module 303 is configured to encode and modulate the transport block 302 onto a time-frequency resource block 306 having a predetermined duration, wherein the time-frequency resource block 306 comprises a plurality of downlink (DL) control and data channels, a guard interval and an uplink (UL) control channel which are organized in a time-division duplex (TDD) manner. The transmission module 307 is configured to transmit a radio signal 308 carrying the time-frequency signal resource block 306 over a radio channel.

The channels may be organized in time as DL control, DL data and UL control, wherein a modulation and coding scheme of a first subframe of a portion of the DL control channel is pre-defined and a modulation and coding scheme of the subsequent subframes is variable. The modulation and coding scheme of the subsequent subframes of this portion of the control channel may be variable according to channel conditions or according to latency requirements. The UL control channel may exist in each frame or may exist once for each multiple frames. The DL control channel of a current frame may comprise information about the modulation and coding schemes and/or a number of symbols used by the control channel of a subsequent frame and/or a modulation and coding scheme of the current or subsequent data channel. The control channels may comprise HARQ and CQI information. The controller 305 may adjust a length of the guard interval based on distance to a base station or on a round-trip propagation delay. The DL data channel may contain more than one TTI.

The third variant of the transmission device 300 is further described below with respect to Figures 1 1 to 17. In a fourth variant of the transmission device 300, the MAC module 301 is configured to generate a transport block 302. The modulation and coding module 303 is configured to encode and modulate the transport block 302 onto time-frequency resources 306, wherein the time-frequency resources 306 comprise at least one symbol in time direction and a plurality of subcarriers in frequency direction. The controller 305 is configured to subdivide the time-frequency resources 306 into a plurality of frequency bands and/or time periods, each frequency band and/or time period corresponding to a certain latency. The transmission module 307 is configured to transmit a radio signal 308 carrying the time- frequency resources 306 over a radio channel.

The plurality of frequency bands may comprise a first latency frequency band, e.g. a first latency frequency band 1801 as described in Fig. 8, a second latency frequency band, e.g. a second latency frequency band 1802 as described in Fig. 8 and a third latency frequency band, e.g. a third latency frequency band 1803 as described in Fig. 8. The second latency frequency band 1802 may comprise a higher number of symbols in time direction than the first latency frequency band 1801 . The third latency frequency band 1803 may comprise a higher number of symbols in time direction than the second latency frequency band 1802.

The controller 305 may assign each frequency band and/or time period to a specific transmission time interval (TTI). The controller 305 may support different TTIs within the time-frequency resources 306 simultaneously. The controller 305 may assign each frequency band and/or time period to a specific user. The controller 305 may determine a TTI assigned to a frequency band and/or time period of a user based on latency requirements, propagation delay and processing delay of a service provided to the user. The transmission module 307 may transmit a signaling message to each user indicating the TTI assigned to the frequency band and/or time period of the user. The controller 305 may group users of similar TTI in a same user class. The controller 305 may determine a scheduling priority factor for a user being assigned to a frequency band and/or time period which belongs to a different user class. The scheduling priority factor may be computed as a function of a ratio of the TTI duration of the different user class to the user's TTI duration. The controller 305 may scatter the resources used for the same TTI throughout the time-frequency resources 306. The controller 305 may continuously modify the ratio of frequency bands or time periods of different TTI classes according to the traffic load of each class of TTIs. The fourth variant of the transmission device 300 is further described below with respect to Figures 18 and 19.

Fig. 4 shows a block diagram illustrating a reception device 400 according to an implementation form. The reception device 400 includes a reception module 401 , a channel estimation filter 403 and a controller 405. In a first variant of the reception device 400, the reception module 401 is configured to receive a radio signal 402 within a transmission time interval (TTI), the radio signal 402 comprising time-frequency resources which comprise a known pilot pattern. The channel estimation filter 403 is configured to estimate the radio channel 406 based on a selection of pilots 404 from the known pilot pattern. The controller 405 is configured to select the pilots 404 from the pilot pattern according to a given latency budget.

The controller 405 may select pilots 404 from current and past symbols of the pilot pattern in the case of a single TTI. The controller 405 may select the pilots 404 from the pilot pattern which minimize a channel estimation error of the channel estimation filter for the given latency budget. The channel estimation error may comprise a mean squared error depending on an auto correlation function of a channel response at the pilots, an auto correlation function of a channel response at all received elements and a cross correlation function of the channel response at the pilots and the channel response at all received elements.

In a second variant of the reception device 400, the reception module 401 is configured to receive a radio signal 402. The radio signal 402 comprises time-frequency resources which comprise a first pilot pattern during a first time interval and a second pilot pattern during a second time interval, wherein the second pilot pattern comprises comb pilots continuously distributed in time direction and preamble pilots continuously distributed in frequency direction. The channel estimation filter 403 is configured to determine a coherence time and a coherence bandwidth of the radio channel based on the second pilot pattern.

The channel estimation filter 403 may determine the coherence time based on a signal-to- noise ratio (SNR) estimation at pilots of the second pilot pattern. The channel estimation filter 403 may determine the coherence time based on a selection of the pilots of the second pilot pattern which have an SNR above a predetermined threshold. The reception device 400 may include a channel state information (CSI) interface, configured to transmit CSI indicating the coherence time and the coherence bandwidth of the radio channel. The first pilot pattern may comprise scattered pilots. The channel estimation filter 403 may estimate channel coefficients of the radio channel based on the first pilot pattern. The reception module 401 may receive signaling information on a subcarrier spacing of the time-frequency resources over a control channel. The channel estimation filter 403 may estimate the channel coefficients of the radio signal 402 depending on the signaled information on the subcarrier spacing. Fig. 5 shows a schematic diagram of time-frequency resources 500 partitioned into a first latency frequency band 501 , a second latency frequency band 502 and a third latency frequency band 503 according to an implementation form.

A transmitter needs to decide upon the TTI window, i.e. the number of symbols per TTI. This choice should be based on several factors such as the latency restriction, the propagation delay, computation power of the receiver and the Modulation and Coding Scheme (MCS).

The timing advance (TA) parameter in LTE, is a control message which is delivered to each user inside a cell. TA informs the user about the time when the user should start transmitting as referenced to the DL time grid. This ensures that all the packets transmitted from all the users in the cell arrive to the base station at the exact same time, so that all UL traffic is synchronized at the base station. Users at the edge of the cell begin their transmission early in order to compensate for the prolonged latency due to the propagation delay.

Link adaptation is a functionality in LTE, where a minimum Quality of Service (QoS) is guaranteed by continuously adapting the Modulation order and coding scheme according to the channel conditions. This means that for bad channel conditions, lower modulation order and coding scheme is used, this implies that the throughput on PHY layer is reduced, hence the MAC layer frame is delivered over several TTIs. In a latency constrained system, this should not be allowed, since a MAC frame arriving later than the latency deadline is considered lost. Hence, according to this disclosure, the MCS is restricted to the latency requirements. For example, low latency applications should have a minimum threshold MCS even if it violates the minimum QoS. For high latency requirement, the MCS may use very low schemes (BPSK or code rates<1/3) in order to make use of the latency tolerance to enhance reliability. Additionally, the pilot overhead should be taken into account when choosing a suitable MCS. This is especially critical when the pilot density is variable.

The total latency including the propagation delay (obtained from TA), is used to determine the number of symbols per TTI. This implies, that the users at the cell edge may have to use fewer symbols per TTI compared to users closer to the base station although they have the same service latency requirement. The total spectrum can be subdivided into different bands each dedicated to a certain latency requirement. Figure 5 illustrates this as an example. The low latency band 501 has a TTI window of a single symbol and may be used for example for V2V (vehicle to vehicle) video applications. The medium latency band 502 has 2 symbols per TTI and may be used for example for voice applications, the high latency band 503 has 3 symbols per TTI and may be used for example for MTC (machine type communication)

thermometers and so on. The key advantage of this segmentation is that a user wishing to change its TTI window can simply configure a software parameter of TTI window length in symbols. However in state-of-the art, the user would have to configure a different subcarrier spacing and new RF parameters.

Fig. 6 shows a block diagram illustrating a transmission device 600 according to an implementation form. The structure of the transmission device 600 may correspond to the transmission device 200 described above with respect to Fig. 2. The transmission device 600 may transmit a radio signal over a radio channel. The transmission device includes a transport block generator 601 , e.g. a media access control (MAC) module as described above with respect to Fig. 3, a channel coding unit 602 and a modulation unit 603, e.g. a modulation and coding module as described above with respect to Fig. 3. The transport block generator 601 generates a transport block and the channel coding and modulation units 602, 603 encode and modulate the transport block onto time-frequency resources 604, e.g. a resource block 100 as described above with respect to Fig. 1 .

Within each TTI, the following information may be present: a data bearing channel Physical Downlink Shared Channel (PDSCH) which carries the main information, control channel Physical Downlink Control Channel (PDCCH) for describing the format of the information transmitted for e.g. MCS, pilots for coherent detection, synchronization sequence. For a single symbol TTI system, the frame structure 604 may look as shown in Fig. 6: The control channel C is centered around the zero subcarrier frequency. By decoding the control channel C the receiver knows the whole occupied bandwidth. Pilot elements R may or may not exist in each symbol depending on the pilot insertion scheme. Synchronization elements exist only at initial synchronization, afterwards synchronization is based on the pilot elements. In this way, the overhead of synchronization elements can be eliminated in order to boost the throughput. In state-of-the art approaches, the synchronization elements are always present, which may present an undesired overhead.

Fig. 7 shows a schematic diagram of time-frequency resources 700 illustrating selected pilots RS and non-selected pilots RN according to an implementation form. In state-of-the art, channel estimation is done within a TTI (1 ms). There are numerous methods to perform channel estimation. However, the channel estimation technique is usually fixed since the TTI is fixed.

In the adaptive TTI frame structure according to the disclosure, the channel estimation filter should be tailored according to the TTI length and channel. For example, a TTI of a single symbol implies that the filter can use only current and past pilots (causal filter). Since the TTI may change continuously, the channel estimation filter must also change continuously.

The complexity of a channel estimation filter is defined by the number of pilots used for channel estimation. Increasing the filter complexity may improve the channel estimation quality; however, it would increase the overall latency due to longer processing time, which contradicts the key target of the low latency transceiver. Hence, in the approach according to the disclosure, a certain latency budget is allocated for channel estimation depending on the computation speed of the receiver. In some cases, the complexity of the channel estimation filter may be the dominant part in the latency budget. For example, a high speed receiver of a vehicle may use 20 pilots to estimate the channel, whereas a simple sensor node may use 3 pilots to estimate the channel for a data location.

As mentioned, the latency budget allocated for processing is mapped to the maximum possible filter complexity. The task is to find which pilots in the vicinity of the data element that should be selected to perform channel estimation. By selecting the best pilots, the minimum channel estimation error is achieved for a given latency budget and receiver complexity.

The Mean Squared Error (MSE) of channel estimation may be defined by the formula:

-^Tr{R _yy - R _yx R^R"} , where y is the channel's response the data elements and x is the channel's response at the pilot elements, and R is the covariance matrix of the signals.

In this disclosure, the pilots which minimize the MSE are selected. This essentially means that according to the channel's fading in time and frequency the pilots are selected. It will be shown below that the transmitter may transmit preamble + comb pilots for a better estimate of the channel's second order statistics. This information can be used to define the adaptive channel estimation window.

In Figure 7 an exemplary pilot selection mechanism is shown. Selected pilots are denoted as RS, non-selected pilots are denotes as RN and considered data elements are denoted as X. The TTI is assumed to be a single symbol, which implies that the channel estimation window can only select pilots from current and past symbols. In this case, the computation speed of the receiver allows for the use of 6 pilots for channel estimation. As shown, more pilots are selected in the time axis compared to frequency axis, which implies that the channel is mostly frequency selective rather than time selective.

As an example, the channel estimation window can change. Assume a device is not moving but located inside a building. In this case the channel is multipath rich, hence the channel is strongly frequency selective. In this case, the channel estimation window does not select a lot of pilots in the frequency direction but rather more in the time direction since the channel does not change due to the stationary user. Once the user starts moving in an open environment with no multipath reflections, then the channel estimation window changes and starts selecting more pilots in frequency compared to time. Fig. 8 shows two exemplary performance diagrams 800a, 800b illustrating mean square error over SNR for different Doppler frequencies.

Fig. 8 shows the results for a simulation of the Symbol Error Rate (SER) 800a and Mean Squared Error (MSE) 800b for the single symbol TTI frame structure compared to the LTE frame structure with the same pilot density and 16 QAM. The channel is ETU with different Doppler shifts. The graph 801 illustrates Doppler shift 5Hz, the graph 802 illustrates Doppler shift 36Hz, the graph 803 illustrates Doppler shift 72Hz, the graph 804 illustrates Doppler shift 108H and the graph 805 illustrates Doppler shift 144Hz. The results are shown in Fig. 8 where solid line is the LTE frame structure and the dashed is the single symbol TTI. Both cases use the same Wiener channel estimation filter order.

Clearly, using causal channel estimation filters implies lower reliability. However, with the proposed channel estimation technique the gap between LTE and single symbol TTI is generally not too wide. Hence, the latency has been shortened by a factor of 14 and the reliability was not severely reduced. Fig. 9 shows a schematic diagram of time-frequency resources 900 partitioned for a High Doppler channel 903 or a high RMS delay channel 905 according to an implementation form. According to the service requirements 902, the available resources 901 in bandwidth 903 and latency deadline 906 can be selected to be used for high Doppler channel 903 where the channel is sampled more frequent in time or for high RMS delay channel 905 where the channel is sampled more frequent in frequency.

In future communication systems, it is expected that some devices will have superior computational power such as vehicles or infrastructure. For those devices, an extended mode is presented in this disclosure, where the transmitter has the freedom to modify the subcarrier spacing within a given time-frequency resource. Note that the motivation of modifying the subcarrier spacing here is intended to enhance the reliability of the link and not to reduce the TTI length as proposed in the literature. For a given pilot density, the distance between pilots in absolute time and frequency changes by changing the subcarrier spacing. Then, the channel estimation performance and hence reliability of the wireless link changes.

The transmitter obtains knowledge about the channel's second order statistics, such as Doppler shift and RMS delay, and tunes the subcarrier spacing according to the channel's condition. For example, for a fast varying channel with high Doppler, the subcarrier spacing is increased. In this way pilots come closer in time and hence channel estimation is improved without increasing the pilot density. Fig. 10 shows a schematic diagram of time-frequency resources 1000 in which different resource blocks 1001 , 1002, 1003 and 1004 carry different pilot patterns according to an implementation form.

In order to obtain an accurate estimate of the channel's second order statistics, the transmitter can change the pilot pattern every fixed period of time to be a preamble and comb pilot (continuous time/frequency) according to resource blocks 1001 , 1004, instead of scattered pilots (as in LTE DL) according to resource blocks 1002, 1003 or preamble only pilot (as in LTE UL). By using the continuous time/frequency pilot structure the receiver can obtain a higher accuracy estimate of the channel's autocorrelation function. The autocorrelation function in time/frequency is compared to reference correlation functions in order to obtain an estimate of the channel's second order statistics which is then fed back to the transmitter. As a byproduct of such pilot pattern, the receiver can obtain a better estimate of the Wiener channel estimation filter coefficients, due to the higher resolution channel's autocorrelation function.

Fig. 1 1 shows a schematic diagram of a TDD radio frame 1 100 having downlink (DL), uplink (UL) and guard interval (Gl) sections according to an implementation form, also referred to as "Alternating Symbol TDD". In LTE, there are different TDD modes where within a TTI (1 ms), the number of DL and UL slots are varied according to the DL to UL traffic ratio. The limitation in this frame structure, is that once several slots are allocated to DL, the feedback from the receiver in the UL arrives after several slots later. Late arrival of a feedback from the receiver increases the Round Trip Time, hence prolongs the latency.

The concept of symbol-wise TTI processing can be translated to TDD systems by having the freedom to alternate between DL and UL in symbols.

In the most basic case, the UL and DL can alternate each symbol. In this case, the Round Trip Time is minimized. However, this frame structure is suitable for situations where the UL and DL traffic is more or less equal (such as V2V or generally D2D scenarios).

The guard interval between the UL and DL symbol should accommodate for the two way propagation delay between the communication nodes. Luckily, in D2D and in most V2V cases, the distances between the communicating nodes are not large. For example, for a maximum supported distance of 1000 meters, the guard interval needs to be 6.67 μβ, which is not quite large, hence the spectral efficiency is not drastically degraded.

In the frame shown in Fig. 1 1 , there are 13 symbols, 7 for DL 1 102 and 6 for UL 1 104. The symbol duration is similar to LTE (71 μβ) and the Gl 1 103 is 6.41 s (supporting a distance of 961 m), so that the total duration of the frame is 1 ms, which makes it compatible to legacy systems LTE.

Furthermore, the ratio between the DL and UL symbols can be tuned according to the ratio between DL and UL traffic. This frame structure may be most suitable for D2D communications which is latency sensitive such as V2V or any short distance communication protocol. Similarly for users which are close to the base station and wish to have a low latency frame structure.

Fig. 12 shows a schematic diagram of a self-contained TDD radio frame 1200 according to an implementation form, also referred to as "Self-contained TDD".

In the concept of self-contained TDD, the TDD frame structure is composed of a DL slot followed by a short UL slot containing ACK information referring to the DL information in the same frame. This way, all the information, control, and feedback is contained in a single packet instead of having the feedback information distributed in another subcarrier band or in arriving after several slots later.

The frame structure in its current form may suffer low spectral efficiencies if not carefully designed. For example, the choice of the Guard Interval should be adequately chosen in order to accommodate for the propagation delay and the processing time at the receiver needed to compute the ACK or NACK.

In this disclosure, the guard interval is minimized as much as possible in order not to degrade the spectral efficiency. Using short guard intervals has two implications: The cell range is shortened; and the receiver computation power must be high to quickly compute ACK within the guard period.

Those two implications are addressed by introducing some modifications to the frame structure as shown in Fig. 12. In the frame 1200 a control field 1201 for TX direction is followed by a data field 1202 for TX direction, a guard period 1203 and an Uplink field 1204 for TX direction.

Fig. 13 shows a schematic diagram of a radio cell 1300 which is subdivided into concentric circles according to an implementation form.

In a new approach, the cell 1300 may be subdivided into concentric circles. In each zone 1302, 1303, 1304, a different guard interval may be chosen according to the maximum expected propagation delay within this zone. For example, in Fig. 13, the guard interval of the users in zone A, 1302 may be shorter than that in zone B, 1303 and C, 1304. This way the frame structure may be adapted to the user distance to the base station 1301 . Note that the total frame duration is fixed. Only the guard interval changes within the frame duration by using a portion of the time from the DL data channel.

Fig. 14 shows a schematic diagram of a radio signal 1400 including a first frame 1401 followed by a second frame 1402 according to an implementation form. The first frame 1401 includes a PDSCH part 1404, a PDCCH part 1405, a guard interval Gl 1406 and an acknowledge part ACK 1407. The second frame 1402 includes a PDSCH part 1408, a PDCCH part 1409, a guard interval Gl 1410 and an acknowledge part ACK 141 1 . Fig. 14 addresses the second implication, which is the assumed receiver computational power. It is not practical to link the guard interval to the computational power of the receiver. Hence the frame structure may be modified to decouple the dependence of the guard interval to the processing power. In the new frame structure shown in Fig. 14, the location of the data bearer channel (PDSCH) is flipped with the control channel (PDCCH) of DL. Additionally, the current control channel of frame N 1401 is assigned to the next frame N+1 1402. This means that after the PDSCH is received, the receiver has the additional time of the control channel and the guard interval to finalize the ACK computation, which should be enough to complete the decoding process. In this way the guard interval becomes solely dependent on the propagation delay, and not on the receiver complexity.

Fig. 15 shows a schematic diagram of a frame 1500 in which the PDSCH 1501 is subdivided into several TTIs according to an implementation form. The frame 1500 has a fixed frame length and includes a PDSCH 1501 , a PDCCH 1507, a Gl 1508 with variable guard interval 1503 and an ACK 1509.

In order to enable low latency communications, the DL slot 1500 can be further subdivided into several TTIs (1 to M) 1503, 1504 and 1505 to 1506. Each TTI can be separately decoded. The UL ACK received may either correspond to each TTI separately, or be a single (N)ACK for all the TTIs (similar to TTI-bundling).

Fig. 16 shows a schematic diagram illustrating an exemplary frame structure 1600 for low latency communication according to an implementation form.

The frame structure 1600 includes a plurality of DL data 1601 , followed by a DL control 1602, a guard interval 1603 and an UL 1604. The total frame duration is 1 ms. There are 14 symbols each having a duration of 66.67 + 4.25 = 70.92μ8, where 4.25μ8 is the CP (cyclic prefix). The guard interval is 6.67μ8. Each symbol can be considered as a separate TTI which contains data and pilots only. The control information for all the TTIs within this frame is contained in the control symbol.

Fig. 17 shows a schematic diagram illustrating an exemplary frame structure 1700 for low latency streaming applications according to an implementation form. The frame 1700 includes a plurality of DL data 1701 and DL control 1702, each of length 1 ms, followed by a guard interval 1703 and an UL 1704.

In certain scenarios, the wireless link should be able to transmit at high throughputs without the need for continuous acknowledgement in the feedback link. For example in video transmission between cars (see-through), if an image frame arrives distorted it is simply dropped or interpolated. Hence, for such streaming applications, the periodicity of the acknowledgements in the feedback may be reduced, so that the UL signal arrives once each several frames as shown in Fig. 17.

Fig. 18 shows a schematic diagram of time-frequency resources 1800 in which time is subdivided into different periods according to a latency aware frequency-segmented resource allocation (LAFRA) approach according to an implementation form, also referred to as "Latency aware resource allocation".

In LTE systems, the smallest unit for allocating resources is a resource block, which is formed of 12 subcarriers <7 OFDM symbols. The base station should allocate the resource blocks to each user and inform the user at which resource blocks he is allowed to transmit. There are various methods to design a Radio Resource Management (RRM) scheme. Usually, the prime goal is to maximize the sum rate, spectral efficiency, while trying to be as fair as possible so that all the users receive a fair portion of the resources. However, the latency restrictions of each user in the cell has not been considered so far in the resource management process.

In this disclosure three main approaches on top of the legacy RRM techniques are presented: The resources can be allocated in symbol-basis, rather than in multiples of 14 symbols. This implies that now the spectrum has a fine granularity in time. Hence, the resource "holes" which could have been present e.g. in the LTE, are now minimized. The spectrum can be subdivided into different bands. In each band, a given TTI window size has a priority over the other TTI windows. For example, a certain band is assigned to users with a TTI of 1 symbol (low latency). In this band it is still possible to have users with longer TTI windows. However, they will have lower priority compared to users with 1 symbol TTI. This approach is also referred to as Latency Aware Frequency-segmented Resource Allocation (LAFRA).

Fig. 18 shows an example of LAFRA, where the wireless system can support a TTI of 1 symbol (Low latency) 1801 , 3 symbols (medium latency) 1802 and 7 symbols (Long latency, LTE) 1803. Hence, the resources are divided into three bands each

corresponding to a TTI window size. The ratio of the three bands is dynamically changed according to the traffic load of each latency requirement. For example, a base station covering a highway, will need to allocate more resources to the low latency band due to the heavy presence of vehicles which need to communicate in a low latency manner.

In another resource allocation technique, the base station subdivides the time into different periods, each corresponding to a certain TTI window. This approach is also referred to as Latency Aware Time-segmented Resource Allocation (LATRA). An example of LATRA is shown in Fig. 19. The base station ranks all the UEs within the cell according to their latency constraints, propagation delay (distance from base station), and processing power. Note that users at the cell edge may be assigned to a different latency band than what they would be assigned if they are closer to the base station due to the longer propagation delay. The base station configures each UE with a suitable TTI window according to the factors mentioned above. The base station then allocates resources in the corresponding band of its TTI. Fig. 19 shows a schematic diagram of time-frequency resources 1900 in which time is subdivided into different periods 1901 , 1902, 1903 according to a latency aware time- segmented resource allocation (LATRA) approach according to an implementation form.

As an example of LATRA, the base station updates the resource allocation each 10 ms, the first 2 ms are allocated for users with 1 symbol TTI 1904, the second 4 ms are allocated for users with 3 symbols TTI 1905, the third 4 ms are allocated for users with 7 symbols TTI 1906, and so on. Within each period, the users are allocated resources according to the channel conditions and the data rate they are expecting. The location of the bands in LAFSR and the location of the periods in LATSR do not have to be adjacently allocated as shown in Fig. 19, but can be randomly scattered throughout the frequency/time grid in order to obtain a certain degree of frequency/time diversity. For both approaches, if the corresponding band/period is fully occupied, then the user is assigned to a different band/period, while having the risk of not being fairly allocated resources compared to other users belonging to this band. Note that the TTI window remains the same for a user even if it is placed in a different band/period. For example, a low latency user with 1 symbol TTI, will still use a 1 symbol TTI even if he is placed in a band/period which uses a 7 symbols TTI for example.

Within each band/period a similar approach to the Modified Largest Weighted Delay First (M-LWDF) scheduling algorithm may be used. For each resource block that is formed of 12 subcarriers ^χ 1 symbol, a priority factor may be calculated for each user. The user with the highest priority factor may be assigned the resource block. The priority factor of user / ^' for the resource block at time m, and subcarrier n may be defined as follows

d _l log(l + )

a

d τ _ι + ε ₀

where γ _ι is the effective SNR at the resource block, r. is the delay tolerance of the service, d _i is the requested data rate of the user, d _l is the average data rate of the user in the past T seconds, ε ₀ is a small fixed number to avoid division by zero, and a° is a TTI privilege factor. The TTI privilege factor depends on the band/period of the user and can be denoted as

Band's TTI length

Own TTI length

This means that when a user migrates to a band/period with higher TTI length, he has higher priority, i.e. a° > 1 . For example, a user with a TTI length of 1 symbol who did not find any available resource at the 1 symbol TTI band/period, migrates to the 3 symbol TTI

3

band/period. Then in this case we have a° =— . Hence, the user has 3 times better chance to get allocated a resource block compared to a user with the same exact channel, data rate, etc. but having a 3 symbol TTI.

It has been shown that LATRA is more suitable for sporadic type of traffic (V2X messages), whereas LAFRA is more suitable for streaming traffic (video). The present disclosure also supports a method for transmitting a radio signal over a radio channel. The method includes: generating a transport block; encoding and modulating the transport block onto time-frequency resources according to a transmission time interval (TTI) window, wherein the time-frequency resources comprises at least one symbol in time direction and a plurality of subcarriers in frequency direction; adjusting the TTI window in symbol units according to a latency budget and transmitting the encoded and modulated transport block of a radio signal carrying the time-frequency resources over a radio channel. The present disclosure also supports a method for receiving a radio signal over a radio channel, the method comprising: receiving a radio signal within a transmission time interval (TTI), the radio signal comprising time-frequency resources which comprise a known pilot pattern; estimating the radio channel based on a selection of pilots from the known pilot pattern; and selecting the pilots from the pilot pattern according to a given latency budget.

The present disclosure also supports a computer program product including computer executable code or computer executable instructions that, when executed, causes at least one computer to execute the performing and computing steps described herein, in particular the steps of the methods described above. Such a computer program product may include a readable non-transitory storage medium storing program code thereon for use by a computer. The program code may perform the performing and computing steps described herein, in particular the methods described above. While a particular feature or aspect of the disclosure may have been disclosed with respect to only one of several implementations, such feature or aspect may be combined with one or more other features or aspects of the other implementations as may be desired and advantageous for any given or particular application. Furthermore, to the extent that the terms "include", "have", "with", or other variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term "comprise". Also, the terms "exemplary", "for example" and "e.g." are merely meant as an example, rather than the best or optimal. The terms "coupled" and "connected", along with derivatives may have been used. It should be understood that these terms may have been used to indicate that two elements cooperate or interact with each other regardless whether they are in direct physical or electrical contact, or they are not in direct contact with each other. Although specific aspects have been illustrated and described herein, it will be

appreciated by those of ordinary skill in the art that a variety of alternate and/or equivalent implementations may be substituted for the specific aspects shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific aspects discussed herein.

Although the elements in the following claims are recited in a particular sequence with corresponding labeling, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those elements, those elements are not necessarily intended to be limited to being implemented in that particular sequence.

Many alternatives, modifications, and variations will be apparent to those skilled in the art in light of the above teachings. Of course, those skilled in the art readily recognize that there are numerous applications of the invention beyond those described herein. While the present invention has been described with reference to one or more particular embodiments, those skilled in the art recognize that many changes may be made thereto without departing from the scope of the present invention. It is therefore to be understood that within the scope of the appended claims and their equivalents, the invention may be practiced otherwise than as specifically described herein.