The Internet is a cooperative message-forwarding system linking computer networks all over the world. Users of the Internet can exchange electronic mail, participate in electronic discussion forums, send files from any computer to any other computer, retrieve information, and even use each others computers directly if they have appropriate passwords. DOWNING ET AL, DICTIONARY OF COMPUTER AND INTERNET TERMS, 239-40 (6th ed. 1998).

Individual users seeking to use the Internet typically gain access via online service providers. The cost of running the Internet is paid primarily by these service providers who, in turn, attempt to fairly distribute those costs among their customers.

In today's market, service providers charge a flat rate for providing Internet access.

This flat rate is either a true flat rate, e. g., a fixed price per month for unlimited access, or it is a flat hourly rate assessed to customers based upon the actual time that they are accessing the Internet.

Since neither of these billing schemes depends upon the actual bandwidth consumed by individual users, the allocation of costs to customers is inefficient and unfair. The following example illustrates this inequity. Two residential users can often times consume vastly different amounts of bandwidth during the time that they access the Internet. A power user may, for example, be downloading graphical or video data, listening to MP3 files, and operating a personal web page. A less savvy user, on the other hand, may use his or her Internet account to send email and to occasionally surf the web. If the service provider for these two users charges them a fixed monthly rate, they will incur the same cost despite having used vastly different system resources. Similarly, if the service provider charges an hourly rate for Internet access and these two users access the Internet for the same amount of time in a given billing cycle, their bills will again be identical even though they consumed vastly different system resources while they were accessing the Internet.

In addition to unfairly allocating costs by failing to account for bandwidth consumption, the present billing schemes used by service providers also overlook fluctuations in system performance. During peak hours, data packets are transported through the various routers, switches, and hubs that form the backbone of the Internet at a much slower rate than during off-peak hours. This fluctuation in system performance is not factored into the standard flat-rate billing schemes presently employed by service providers. The present invention overcomes these drawbacks by providing service providers with a more flexible billing framework, which accounts for actual bandwidth consumption.

Summary of the Invention In one aspect, the invention comprises a method for billing users for Internet access based upon the actual bandwidth consumed by the user while he or she is accessing the Internet. The method can be practiced by executing a software package referred to herein as the billing rule. According to this aspect of the invention, a service provider is capable of specifying billing criteria, which is then used by the billing rule to determine if and how much a user should be billed for accessing the Internet. The service provider could, for example, bill users based upon the time of day that they are accessing the Internet, charging higher rates for peak-use periods. In addition, service providers can monitor the total erroneous data bytes transmitted to or from a network user's device. In this way, the service provider could reduce a user's bill if he or she encountered an unusually high amount of erroneous data transmissions during a particular billing cycle.

In another aspect, the invention comprises a system that implements the billing rule software. This aspect of the invention is comprised of an Internet Protocol network, a hardware operating platform, the billing rule software, and an interface between the billing rule software and the hardware operating platform. The system is capable of communicating with network user's devices in order to determine their bandwidth consumption. This communication is facilitated by executing the billing rule software. The billing rule software can be stored in the internal memory of the hardware operating platform.

Brief Description of the Drawing The invention is described with reference to the several figures of the drawing, in which, Figure 1 is a flow chart illustrating a method for billing a network user for accessing the Internet; Figure 2 is a flow chart showing the creation of a device object; Figure 3 is a flow chart depicting the functions performed by the class of methods comprising the billing rule; and Figure 4 is a block diagram showing a system for billing network users for Internet access based upon bandwidth consumption.

Detailed Description The invention disclosed herein provides an alternative method and system by which service providers can bill network users for providing them with access to a network. The types of service providers who may find this invention useful include those providing network access via telephone lines, cable, or through the airwaves. In one aspect of the invention, a method that allows service providers to bill network users for accessing the Internet is discussed. Unlike many prior art billing schemes which bill network users at a flat rate, network users in a preferred embodiment of the present invention are billed according to their actual bandwidth consumption. The billing rule disclosed herein may, for example, be a Java class of methods that can be run by a service provider.

In a preferred embodiment of the billing rule, a network user is billed for accessing the Internet according to the steps shown in Figure 1. According to this embodiment, a service provider must first create 10 a file of type. Java. The. Java file is then compiled 20 into a file of type. class with a Java compiler. The. class file is then loaded into a polling engine where it will eventually be used by the polling engine to determine what criteria should be used when billing network users.

According to this embodiment, the service provider could use a Microsoft Windows- type management interface when creating 10 the Java class file containing operational parameters. The operational parameters specify which data counters the service provider would like the polling engine to base its billing calculations on during

execution of the billing rule. The operational parameters can be chosen from the following list of data counters: if InOctets-number of bytes received iflnUcastPkts-number of unicast packets received if InNUcastPkts-number of non-unicast packets received iflnDiscards-number of inbound packets discarded if InErrors-number of packets received in error iflnUnknownProtos-number of packets received with unknown protocol type ifOutOctets-number of bytes sent ifOutUcastPkts-number of unicast packets sent * NUcastPkts-number of non-unicast packets sent ifOutDiscards-number of outbound packets discarded ifOutErrors-number of errors on outbound packets The data counters are described more fully in the Internet Engineering Task Force's Management Information Base for Network Management of TCP/IP-Based Internets : MIB-11, RFC 1213, the contents of which are incorporated herein by reference. As can be seen from the above listing, each data counter represents an aspect of the bandwidth consumption of a network user's device. In accordance with this invention, the data counters are retrieved by the polling engine. The polling engine accumulates the results of polling and passes these results to the billing rule at the end of the polling interval.

When the service provider creates 10 the Java class file, he or she chooses which of these data counters will be collected and stored by the polling engine. These choices become the operational parameters and are stored in a Java class file. In an additional embodiment of the present invention, the service provider could alter the cost of accessing the Internet based upon such factors as time of day that the network user accessed the Internet, whether the total bytes of data transmitted and received during a particular session exceeded a threshold value, or whether the entity accessing the network is an individual or a business. If the service provide chose to structure billing in this way, it would include these criteria in the Java class file.

In a preferred embodiment, a polling engine executes the billing rule software.

The billing rule software could be a Java class of methods. As part of its duties in executing the billing rule, the polling engine polls each network device that is accessing the Internet. In these polling exchanges, the polling engine accumulates data for each of the operational parameters specified by the service provider. Thus, in a preferred embodiment, a service provider can create a Java class file directing the polling engine to accumulate data for as many of the counters listed above as the service provider wishes to monitor. In addition, service providers can create many different Java class files, enabling the polling engine to monitor different counter data for different network users. After the service provider compiles 20 the Java class file, it is stored 30 in a data storage unit. The data storage unit could be a database, computer hard drive, or the like.

Once the Java class file has been stored 30, some of the operational parameters, comprising such information as an IP address and a billing rule name associated with a particular network user, are provided 40 to the software via a web browser or a text file. The polling engine uses this IP address and billing rule name to create 50 the device object. Each device object created 50 by the polling engine represents a unique device configured to operate according to this aspect of the invention. Since it is likely that a plurality of devices will be accessing the network according to the invention disclosed herein, it may be necessary for the polling engine to create a plurality of device objects, each of which is used to distinguish one device from the next. Once a first device object corresponding to a first network device has been created, the first device object uses a first billing rule name passed from the web browser or text file to load a Java class file containing the billing rule byte code. At this point, additional operational parameters, such as the polling interval and the polling period each discussed further below, are extracted from the billing rule. After this step is complete, the device object has been constructed and may comprise: (1) a first device's IP address; (2) a first Simple Network Management Protocol ("SNMP") port number; (3) a first billing rule name; and (4) a first billing rule parameter.

Figure 2 shows a detailed depiction of how the polling engine creates 50 the device object. As a first step in this, the first device object dynamically loads 51 the first billing rule into a Java virtual machine. After the first billing rule has been

loaded, the first device object creates 52 an instance of the first billing rule. At this point, the Java runtime constructor invokes 53 a first billing rule default constructor.

As can be seen from Figure 2, the next step entails the initialize method of the first billing rule being called. When the initialize method of the first billing rule is called, a billing rule parameter is passed 54 as the sole parameter. The first device object then allocates 55 the accumulator array, which is where the polling engine will eventually store the results it obtains. The final step in this embodiment comprises the first device object creating 56 the necessary structures for the SNMP polling of the first device. In addition, the first device object maintains a copy of the operational parameter indicating the start of the polling interval. Once the first device object has initialized the first billing rule, the polling engine places the first device object on a polling list. In a preferred embodiment, the polling engine continuously loops through the polling list, sequentially polling each of a plurality of devices.

After the first device object has been created, it is possible for the polling engine to call the remaining methods in the billing rule. These methods, and the order in which they are called, are depicted in Figure 3. As can be seen from Figure 3, the first step performed during execution of the billing rule is an initialization 110. In this embodiment, the first device object is used to initialize a first billing rule. After initialization has been completed for the first device, the polling engine determines 120 the polling interval and the polling period from the operational parameters in the first Java class file. The begin time for the polling interval is a Gregorian Calendar object, which is part of the standard Java Application Programming Interface. The frequency at which polling will transpire during the polling interval is referred to as the polling period and may be given in seconds. The counters that can be polled are chosen from the list set forth above with reference to the MIB-11, RFC 1213 standard.

At the end of the polling interval, the results that were accumulated are delivered to the polling engine.

After these preliminary operational parameters have been extracted from the first Java class file, the polling engine begins polling the first device and accumulating 130 values for the counters specified by the service provider. During each poll, the polling engine determines 140 if the polling interval has ended by comparing the current time to the end of the polling interval as specified by the billing rule. The

method used to determine if polling interval has ended is also a Gregorian Calendar object. Once the polling interval has ended, the polling engine stops accumulating counter data, calls a method in the billing rule for generating usage billing data, passes the accumulator array, and sends 150 a data array of the specified counters to the billing rule so that the billing rule can compute the usage billing data for the first device. The data array contains the accumulated counters gathered during the polling interval.

When the billing rule finishes computing the usage billing for the first device, it passes one of two values back to the polling engine. The billing rule returns either a null value, indicating a non-billable event, or it returns data constituting a billable event. If the value returned is null, the polling engine resets 180 the billing rule. If, on the other hand, a billable event is returned, the polling engine may instantiate and pass 170 a billable event object downstream. The billable event object could be passed downstream to another component via an interface. In one embodiment, this interface could be a CORBA device. In another embodiment, the polling engine may simply pass the billable event object downstream across a network to a computer component, which in turn determines the appropriate action to take with respect to the billable event object. An example of an appropriate action might be to reformat the event and pass it on to a billing system maintained by the service provider, thereby placing a charge on the network user's account.

Distinguishing a billable event from a non-billable event can depend on a number of factors determined ahead of time by the service provider when it establishes the operational parameters in the Java class file. The operational parameters described here are actually an arbitrarily complex piece of Java code. For example, the service provider may stipulate that it will not generate a billable event for Internet access if the network user consumed less than a threshold value of bandwidth while accessing the Internet. Alternatively, the service provider could determine that it will not bill network users for Internet access of less than a threshold value of minutes, or for use occurring during a particularly low usage period. Because the polling interval and the polling period are retrieved from the billing rule at runtime, the values returned by the billing rule from one polling interval to the next

are computed using Java code, providing the operator with maximum flexibility in determining their respective values.

Referring again to Figure 3, resetting 180 the polling engine clears the counter arrays and prepares the polling engine for the next polling interval. After the polling engine has been reset, it determines 120 when the next polling interval should begin and the polling period, which is the frequency at which the network user's device will be polled.

In an additional embodiment, the polling engine can maintain accumulated values for the various data counters polled on each network device as 64-bit values.

Moreover, the polling engine can subtract an internally computed amount of overhead associated with polling each device so that network users are not charged for being connected to the network during the time that the polling engine is polling a particular device.

In a preferred embodiment shown in Figure 4, a system for billing network users could be implemented on an Internet Protocol, commonly known as an IP, network 310. The IP network in a preferred embodiment may further comprise a web browser. One function of the web browser could be to allow input of the parameters for polling a single network user's device and to otherwise manage the system.

The hardware operating platform 320 in a preferred embodiment may be comprised of a computer, a processor 330, and interface 335, two Local Area Network cards, 340 and 345, respectively, and a data storage unit 350. Once the service provider configures the hardware operating platform 320, he or she may load the billing rule software into the running system from a predefined location in the installation tree. In this way, the hardware operating platform 320 can be used to store the billing rule software. In a preferred embodiment, the service provider can remotely access the billing rule software via the Internet. Additionally, a text file comprising a listing of the plurality of network user's devices 360, which are configured to operate on a usage basis, can also be stored in the data storage unit 350.

Other embodiments of the invention will be apparent to those skilled in the art from a consideration of the specification or practice of the invention disclosed herein.

It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.

What is claimed is: