Cellular telephone systems rely on a web of base stations that receive and transmit the cellular signals to users. A major cost of cellular telephony is due to the number of base stations that are required to cover a given area. Reducing the cost of cellular telephony requires reducing the number of base stations per unit area, which in turn requires either increasing the range of the users'antenna or increasing the range of the base stations'antenna.

One solution to increasing the range of the base stations is through the use of so-called"smart"antennas, which transmit a narrow, focused beam to each user. These smart or adaptive antennas have the potential to provide larger area coverage, reduced signal outages, and improved frequency reuse factors. In this area, prior work has been done on adaptive antenna algorithms and fixed switched beam techniques to improve the uplink performance of existing and future base stations.

Field tests have been conducted to compare the uplink performance of smart adaptive antennas to conventional sector antennas, fixed multibeam antennas and other candidate antenna systems for range extension and interference suppression. Significant results were obtained in the field trial

showing that a 40% reduction in the number of base stations can be obtained and capacity can be more than doubled even with rapid fading with fully adaptive uplink antennas with four antenna elements.

Experimental work to this point has focused on the uplink channel (mobile-to-base) which is a key determinant for many system design parameters, such as coverage, mobile power consumption, and capacity. The uplink channel (mobile-to-base) requires greater improvement than the downlink channel (base- to-mobile) since many cellular and proposed 1.9 GHz Personal Communications Services (PCS) systems are uplink limited due to the disparity in mobile and base station transmit powers. When low power portable phones are accommodated on the uplink channel, increasing the power output of the user's telephone is not possible because of regulatory restrictions and battery life considerations. As has been demonstrated, smart adaptive antennas can be used to balance the link by improving uplink signal quality and provide additional gain to increase the range, but additional power is needed on the downlink channel to compensate for path loss at larger ranges.

To balance the links and compensate for up to an additional 9 dB of path loss at 1.9 GHz PCS versus cellular frequencies, additional transmit power and/or higher transmit antenna gain is needed at the base station. Smart antennas can be used to provide a directive downlink which steers energy in the direction of a desired mobile and reduces interference to other users. This will permit overall lower transmit power levels for equal coverage or enhanced coverage with power levels equal to a conventional system. However, two issues complicate the downlink strategy for smart antennas in the North American Digital Mobile Radio IS-54/IS136 TDMA system: (1) the use of separate frequency bands for the uplink and downlink, and (2) the requirement for a continuous downlink for all users in a frequency channel.

In a frequency division duplex system, such as TDMA, the RF propagation environment differs at the transmit and receive frequencies and is

time varying. Therefore transmitting with the same antenna pattern as that used for reception will result in nonreciprocal transmit and receive channels and it is unlikely that adequate downlink performance will be obtained. For this technique to work properly, either the transmit and receive frequencies must be within the coherence bandwidth, or in slow fading conditions time division retransmission (different time slots in the same channel are used for receiving and transmit) must be used. Neither of these conditions are met in TDMA.

IS-54/IS136 TDMA systems require a continuous downlink to all (up to 3) users in any occupied frequency channel. The downlink signal must be maintained at a constant level for the duration of the frame. The continuous downlink allows the mobiles to use the adjacent time slot signals for gain setting, equalizer training, and synchronization. This downlink requirement further complicates the smart antenna downlink strategy. Forming a single beam as described above or selecting a fixed beam with a multibeam antenna for the desired user may not provide a continuous downlink to all users thereby potentially degrading mobile performance.

The present invention is therefore directed to the problem of developing a downlink technique that overcomes the above obstacles without requiring changes to the TDMA air interface standard or to the mobiles.

SUMMARY OF THE INVENTION The present invention solves this problem by using multiple beams and power control to optimize the downlink antenna pattern and power transmitted to each user. As a result, greater coverage is obtained and less co-channel interference is experienced.

According to the present invention, in a wireless communication system, a method for communicating between a base station and a plurality of users includes using the same antenna pattern for transmission to all users, and adjusting the antenna pattern to improve a performance measure of the system.

In this case, the performance measure can include the bit error rate, the received signal strength, and/or co-channel interference measurements.

According to the present invention, another method for communicating with wireless users comprises the steps of establishing a power level in a transmit antenna beam from a base station based on signal quality measurements of users in the transmit antenna beam, and adjusting the power level for the transmit antenna in accordance with changes in the signal quality measurement of the user in the transmit antenna beam whose path loss is greatest. In this case, if a new user enters the transmit antenna beam, the power level is adjusted in accordance with the new user's signal quality measurement if the new user has a greater path loss than any existing users.

When establishing an initial power level for a new user, according to the present invention, the initial power level for each selected beam is set at a maximum power level for that beam. In the alternative, the initial power level can be estimated based on a received signal strength at the base station.

Yet another method of the present invention includes the steps of using a fixed multibeam antenna array on the downlink, maintaining a continuous downlink to all users in a channel with a minimum required transmit power in that channel, determining in which of a plurality of transmitting antenna beams each wireless user is located, and turning the selected beams on continuously at a constant power level for the entire frame during transmission. As with the above methods of the present invention, this method can also include the step of measuring a downlink bit error rate and received signal strength at each user, reporting the downlink bit error rate and received signal strength for each user to the base station, and using the downlink bit error rate and received signal strength to adjust downlink power. In this case, the transmit power for each beam is adjusted slowly, e. g., not more than once in a frame in 1 dB steps.

Hysteresis can be added to avoid rapid changes in the transmitted power level.

According to the present invention, an apparatus for communicating with a wireless user includes an uplink receiving antenna, a radio unit, a beam scanning receiver, a downlink antenna, a splitter, a power controller, and a duplexer. The uplink antenna receives the signal from the wireless user. The radio unit is coupled to the uplink antenna, and includes a receiver receiving the signal from the wireless user, and a transmitter outputting a signal for the wireless user. The downlink antenna has a plurality of antenna beams. The splitter is coupled to the transmitter and splits the signal into a plurality of signals, one for each of the plurality of antenna beams. The power controller is coupled to the splitter and provides independent attenuation for each of the plurality of signals. The duplexer is coupled to the power controller, the beam scanning receiver, and the downlink antenna. It receives the plurality of signals from the power controller and outputs them to the downlink antenna. The duplexer also receives a plurality of signals from the downlink antenna and outputs them to the beam scanning receiver. The beam scanning receiver determines in which of the plurality of antenna beams the wireless user is located. The power controller adjusts the output power for each of the plurality of signals based on data included in the signal received from the wireless user.

BRIEF DESCRIPTION OF THE DRAWINGS FIG 1 depicts a downlink scenario with fixed beams and power control according to the present invention.

FIG 2 depicts predicted coverage versus the number of beams.

FIG 3 depicts an example of a TDMA smart antenna system according to the present invention.

DETAILED DESCRIPTION The present invention provides a technique for using downlink smart antennas at the base station to increase gain and improve directivity in the

downlink channel thereby improving mobile signal quality, mitigating downlink multipath, improving frequency reuse, and reducing interference to other users.

The smart antenna downlink technique of the present invention does not require changes to the IS-54/IS-136 Time Division Multiple Access (TDMA) air interference standard or to the mobiles. (Note: As used herein, TDMA refers to the IS-54/IS-136 digital TDMA standard, which is the same for both cellular and PCS, although the frequencies differ.) The uplink and downlink benefits of these smart antenna techniques of the present invention can be obtained with hardware and software modifications confined to the base station.

Shown in FIG 1 is a base station 11 with three sectors each with downlink antennas 12,13,14. Antenna 14 has beams four beams where beams 1 and 3 reach mobile 3, and mobiles 1 and 2, respectively.

An example downlink scenario of the present invention is shown in FIG 1, in which three users are on the same frequency channel and two of the four beams are turned on for transmission. According to the present invention, mobile 1 controls the power level in beam 3 because it has greater path loss than mobile 2. Less power is needed in beam 1 because of the proximity of mobile 3 to the base station. Adding hysteresis to the beam selection process or increasing the switch decision time reduces ping-ponging and increases the average dwell time on a given beam. The initial power level for each selected beam for new mobiles will be at the maximum power level for that sector or beam, but other initialization procedures could also be used, such as estimating the power level for the base station transmitter based on the Received Signal Strength Indicator (RSSI) at the base station.

A technique according to the present invention uses a fixed multibeam antenna array on the downlink to provide increased gain to the desired user and maintain a continuous downlink to other users in the channel with the minimum required transmit power. A scanning receiver is used to select the beams in which the mobiles are located. During transmission the selected beams are

turned on continuously at a constant power level for the entire frame. TDMA supports mobile measurements of the downlink bit error rate (BER) and RSSI to support mobile assisted handoffs. A key feature of this technique is the use of these measurements to perform downlink power control. The BER and RSSI reported from each mobile to the base station along with the RSSI measured at the base station are used to control the power on each downlink beam to minimize the transmit power and reduce interference levels. Separate power control is maintained for each beam based on the active mobile or mobiles in that beam. The transmit power for each beam is adjusted slowly so that the downlink appears continuous to the mobiles. Hysteresis is added to avoid rapid changes in the transmitted power level. Several power control algorithms may be used. One approach tracks the mobile reported RSSI to adjust downlink power and uses a BER based threshold.

A second approach tracks the signal strength received at the base station to adjust the downlink power and also uses a BER-based threshold. The second approach has the advantage of less delay in the signal strength measurement information at the base station as compared to that from the mobile. This received signal strength can be that received by the scanning receiver, or that of the dual-diversity receiving antennas for the mobile signal. The former signal strength measurement has the advantage that the transmit antenna pattern (which is the same as the receive antenna pattern) is included in the received signal strength measurement, while the latter technique uses dual diversity to average the fast Rayleigh fading better to obtain a more accurate measure of the shadow fading.

A comparison of the coverage gains of three downlink approaches using a fixed multibeam antenna with M non-overlapping beams is shown in FIG 2.

Assuming that the total amount of power transmitted from the base station is fixed, this plot shows coverage gain over a single beam system versus the number of beams for three approaches: (1) discontinuous downlink, (2)

continuous downlink with ideal power control, and (3) continuous downlink without power control. With a discontinuous downlink, the total amount of power allowed can be transmitted on a single beam which can be switched between time slots. A discontinuous downlink provides the greatest coverage gain, but this is not possible with TDMA systems without a change in the standard. For a continuous downlink, without downlink power control, the power amplifier limits reduce coverage gains with fixed switched beams. Also, this approach may not provide sufficient interference reduction to improve capacity.

A continuous downlink with fixed switched beams with power control as described in the present invention can increase coverage while complying with the TDMA standard.

With this approach a fully smart antenna system for TDMA could consist of an adaptive uplink system with multiple antennas and a fixed switched beam downlink system with power control. The uplink and downlink systems are independent. Downlink beam selection is based on uplink beam selection with power control for each beam based on the mobiles in that beam. The TDMA smart antenna system described here is shown in FIG 3. The uplink could use dual polarized or conventional collinear antennas. Each radio has up to four receivers and a separate beam scanning receiver is shared among several radios.

With this configuration signal quality is improved and a frequency reuse of 3 or 4 may be possible thereby doubling the system capacity.

Turning to FIG 3, a pair of antennas 41 receive a transmission from the users and pass them to an adaptive receiver 42 in the radio unit 43 of the base station 44. The transmitter 45 is coupled to a splitter 46 which splits the signal into four signals, each of which is attenuated by attenuators in the power controller 47. The signals then pass through amplifiers 48, which in turn pass them to the duplexers 49.

The duplexers 49 receive and send signals to the multibeam antenna 50.

Signals that are received from the multibeam antenna 50 by the duplexers 49 are passed to the switch 51, which selects one signal to be passed to the beam scanning receiver 52. There is one beam scanning receiver 52 for each N radios.