BACKGROUND OF THE INVENTION As the use of wireless communications network had grown, so have the number and type of services and features available to the"mobile"or wireless user. Such services and features have placed additional demands on the network resources and the service providers which operate and maintain the network. For example, there were more than 50 million mobile phone subscribers in the United States alone at the end of 1997, and the number of mobile phone subscribers is growing at an annual rate of about 50%. In addition, the growing number and availability of mobile station features (such as caller ID, call block, and mobile phone location services) place heavier processing burdens on existing networks.

Service providers are finding it challenging to keep pace with these increasing network demands. Accordingly, as more subscribers simultaneously make calls and use more mobile station features, networks become congested. A result of network congestion is that calls to and from mobile stations do not go through, and existing calls risk disconnection. A consequence of call failures and disconnections is that mobile subscribers are left frustrated, and may change service providers.

To lower the probability of service disruptions, service providers are

increasing network capacity. One way of growing network capacity is by purchasing more network hardware from an existing vendor. However, new network hardware that is compatible with existing network architecture may not exist, or may be prohibitively expensive. Furthermore, network hardware from a new vendor is often difficult to integrate with existing network hardware. To overcome the issues of cost, compatibility, and provide for future network growth, service providers are utilizing Internet Protocol (IP) networks as the transport mechanism to permit growth in capacity and to accommodate future network demands.

An IP-based network can provide a multi-purpose transport mechanism for all sorts of data types including voice, video or data (collectively called"traffic"). Moreover, the widespread use of IP makes it an ideal platform for integration of other communication platforms, such as wireless and wireline or public networks. For example, a base station interface could be configured as the gateway or interface between other components of the wireless network and an IP network. Likewise, a public network interface could provide the interface mechanism between the IP network and data originating in a public network such as the Public Switched Telephone Network (PSTN). Furthermore, within the IP network, a subnet functions as an exchange mechanism between the base station interface and the public network interface. To monitor traffic, establish new connections, or place addition features on the network, for example, the IP network typically utilizes an Operations and Maintenance (O&M) function. Other functions, such as call placement, handshaking, connection supervision and monitoring are accomplished through a system control interface.

In operation, when a voice call is made from a mobile station in the wireless network, the call reaches a base station via radio waves, which, in turn, are routed to the base station interface for formatting into data packets suitable for transport over the IP network. The call enters one of a plurality of subnets and is directed to the public network interface leading to other

elements of the PSTN. Calls may be routed to a mobile phone from a public network in a similar manner.

An advantage of IP networks is that they provide widely understood, predefined standards for transporting many information types including voice, video and data. IP is said to be an"open" (meaning non-proprietary) standard. Furthermore, since IP networks are non-proprietary, then benefit from a large number of installed hardware platforms. For these reasons, IP networks are becoming a popular choice for expanding existing network computing.

A disadvantage of IP networks is that they provide a"connectionless unacknowledged network service."IP networks"send and forget"each packet individually across the network nodes and the data packet is re- assemble at a destination point. During times of heavy traffic and network congestion, packets are sometimes lost, corrupted, or arrive out of order.

The result is that voice, video and data reception quality is low during periods of high traffic. Data integrity is jeopardized whenever 100% of the packets don't arrive 100% intact. For some minimal level of voice packet or video packet transfer failures are acceptable; in other applications, however, even minor deviations in packet integrity result in complete data loss and/or a request for retransmission.

While a retransmission algorithm can then be used to recover lost packets, the increased overhead results in more network congestion and decreased capacity for the network overall. When used as a transport mechanism between a wireless and public network data is introduced at a subnet of the IP network.

Often, multiple signaling channels are available for packet transport over the IP portion. By default, however, many IP interfaces introduce data at a predetermined subnet regardless of traffic conditions.

Therefore, what is needed is a method and system for managing the transfer of voice, video, and data through an IP network that avoids network congestion. A solution that gathers traffic data regarding a network to determine a best route for carrying voice, data or video between a base station interface and a public network interface in a reconfigured IP network would provide numerous advantages.

SUMMARY OF THE INVENTION The present invention provides an IP network configuration which utilizes multiple subnets to achieve load leveling. A system control interface analyzes network congestion information based on traffic data gathered by an operations and maintenance node and other data gathering devices, and, based on the indicated congestion conditions, a preferred subnet is selected as a pathway for all or part of an existing data stream or as a pathway for a new data stream.

According to one embodiment, disclosed is a method of controlling traffic over multiple IP subnets. The method comprises the steps of gathering traffic data associated with the multiple IP subnets, and analyzing the traffic data to determine which of the multiple IP subnets provides the most appropriate connection for transmitting a data stream between the wireless network and the public network for achieving a desired traffic distribution over the multiple IP subnets.

The method may include the steps of introducing a data stream onto a first subnet, and then directing a new data stream (or redirect an existing data stream) through a second subnet. The step of analyzing may use statistical decision theory to analyze traffic data when selecting a preferred subnet.

Furthermore, the data stream may include voice packets, video packets, or data packets.

A data stream may flow across more than two subnets of the IP network. Accordingly, a first data stream portion is directed between the base station interface and the public network via a first subnet, and a second data stream portion is directed between the base station interface and the public network interface via a second subnet, and so on.

The method of the present invention can be applied where a voice data stream initiated by a phone call consists of voice (or audio) packets introduced into the IP network. Thus, several phone calls can be directed across a first subnet as a plurality of voice data streams. Then, some or all these voice data streams, or a voice data stream initiated by a new call, may be routed between the base station interface and the public network interface through a second subnet.

In another embodiment, disclosed is a method of relieving IP network congestion in a communications system. Comprising an IP network for carrying traffic between a base station interface and a public network interface. The method comprises the steps of gathering data associated with a first subnet and a second subnet, and analyzing the data in order to select either the first subnet or the second subnet as a preferred subnet. Then, the method directs traffic (additional traffic or existing traffic) between the base station interface and the public network via the preferred subnet.

Also disclosed is a system for controlling traffic over multiple IP subnets (control system) switchable for use in connection with a communications system comprising a public network, a wireless network, and an IP network configured as a transport mechanism between the public network and the wireless network. A public network interface is configured as a gateway between the IP network and the public network, and a base station interface functions as the gateway between the IP network and the wireless network. The control system comprises at least two subnets in communication with the base station interface, an operation and maintenance

interface for gathering traffic data associated with the multiple IP subnets, and a system control interface for analyzing the traffic data to determine which of the multiple IP subnets provides the most appropriate connection for transmitting the data stream between the wireless network and the public network for achieving a desired amount of traffic distribution over the multiple IP subnets.

BRIEF DESCRIPTION OF THE DRAWINGS Other aspects of the invention, including specific embodiments, are understood by reference to the following detailed description taken in conjunction with the appended drawings in which: Figure 1 shows a communications system that employs an IP network to send data packets, voice packets and video packets between a public network and a wireless network; Figure 2 is a communications system having an IP network with two subnets configured according to the teachings of the present invention; Figure 3 illustrates an algorithm for determining whether an IP network carries voice traffic; Figure 4 is a process flow diagram illustrating a method of placing a new call on an IP network which is not currently carrying voice data; Figure 5 is a process flow diagram illustrating a method of placing a new call on a communications system which is currently carrying voice data; and Figure 6 is a block flow diagram of a method of load leveling and reducing traffic congestion according to one embodiment of the invention.

References in the detailed description correspond to like references in the figures unless otherwise indicated.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS The present invention provides a method and system for analyzing traffic levels in an IP network to select an efficient signaling channel for the transport of data over the IP network. In particular, data collection devices, such as an operations and maintenance (O&M) function, gather network statistics, data or other information relevant to IP network congestion (collectively,"traffic data"). Based on an analysis of the traffic data, a system control interface selects a preferred subnet as a pathway for all or part of an existing data stream, or a new data stream. Since traffic may be directed across multiple subnets, the present invention can be used to relieve congestion in a congested subnet and to achieve load leveling (meaning approximately even traffic levels) across multiple subnets. While the present invention is described particularly in terms of two subnets carrying a voice data initiated by a phone call, it should be understood that the principes disclosed herein in other communication systems where multiple entry points into the system are available.

First, a brief overview of the Internet Protocol (IP) and conventional IP addressing schemes. IP is one of the protocols in the Transmission Control Protocol/Internet Protocol (TCP/IP) suite. The IP protocol provides a standard way for users to send and receive (route) traffic. IP takes raw voice, data or video and, when necessary, breaks it down into manageable units, called datagrams. The datagrams are transmitted within the IP network in standardized traffic units called packets. Packets are typically created by electronically attaching a set of electronic data, known as a frame header, directly to the datagram. The frame header will generally contain information such as a source address of the packet, a destination address of the packet, and other information, such as a protocol number and a checksum, for

example. Accordingly, the IP protocol is responsible for routing each individual packet to a destination address.

IP addresses typically comprise a 32 bit address which may look like "128.12.5.155". For example, assume that this is the network address for a mobile station. In this example, 128.12 identifies the IP network where the mobile station is located, and. 5. articulates the subnet associated with the mobile station. Finally,. 155 specifically identifies the mobile station. Thus, all datagrams destined for the mobile station will have a frame header comprising the destination address 128.5.12.155. Likewise, all datagrams generated at the mobile station 10 will have a frame header comprising an origination address of 128.5.12.155.

With reference to Figure 1 therein is shown a communications system that employs an IP network 5 to send data, voice and video (collectively, "traffic") between a public network 70 and a wireless or mobile network 7. In figure 1, the wireless network 7 is shown to comprise a mobile station 10 within a geographic service area known as a cell 30. The cell 30 in radio communication with a base station 20 which manages call activity from all mobile station device within the cell 30. The mobile network 7 could be based on the Global System for Mobile Communications (GSM), the Advanced Mobile Phone System (AMPS), the Pacific Digital Cellular (PDC) network, or another similar standard adopted for use in connection with Public Land Mobile Networks (PLMN) mobile network, for example.

The base station interface 40 functions as a gateway to enable communication between the IP network 5 and the mobile network 7 by converting traffic received from the base station 20 (and perhaps partially processed by the base station 20) into a format compatible with the IP network 5. This means that the base station interface 40 breaks up the traffic into datagrams (if necessary), frames the datagrams to form packets, and directs the transmission of the packets through the IP network 5 via a subnet

50 to the public network interface (or gateway) 60. Similarly, the base station interface 40 may receive IP formatted traffic destined for the mobile station 10 through the subnet 50. In this situation, the base station interface 40 converts the traffic into a format compatible with the requirements of the mobile network 7. Typically, the base station interface 40 processes outgoing traffic from several base stations across wireless or wireline communications channels 22.

The public network interface 60 functions as a gateway to public networks 70 such as Public Switched Telephone Network (PSTN) 72, Public Land Mobile Network (PLMN) 74, and Private Branch Exchange/Primary Rate Interface (PBX/PRI) 76, for example, so that a call may be routed into, or out of, the IP network 5. Accordingly, as traffic arrives from a public network 70 to the IP network 5, the public network interface 60 may break up the traffic into datagrams, frame the datagrams, and direct the routing of the packets through the IP network 5 to an appropriate base station interface. Likewise, the public network interface 60 processes traffic destined for a public network 70 from an IP format into an appropriate public network format. For example, the IP packetized data may be modulated with an analog carrier signal and transmitted along the facilities of the Public Switched Telephone network (PSTN) such as twisted pair copper wiring. Thus, the base station interface 40 and the public network interface 60 function similarly, but at opposite ends of the IP network 5.

A service provider or network manager may partition the IP network 5 so as to create two or more subnets 50,55. For example, the network manager may subdivide the IP network 5 into a first subnet 50 and a second subnet 55. The subnets 50,55 could, for example, each service a different sector or geographical area within the service area of the IP network 5. As another example, the network manager could organize the IP network 5 by assigning a subnet to each base station interface 40. This structure is illustrated in figure 1 where the first subnet 50 serves a first base station

interface 40 and a second subnet 55 serves a second base station interface 52 though, not shown, it should be understood that the second base station interface 42 is connected via communication channel 22 to a mobile network similar to the mobile network 7. Of course, other organizations of subnets 50, 55 are also implemented. In essence, the first subnet 50 is seen as serving as a transport mechanism between the public network interface 60 and the base station interface 40 associated with cell 30. Likewise, the second subnet 55 serves as a transport mechanism between the public network interface 60 and a second base station interface 42 associated with other cells (not shown).

Figure 2 is a communications system having an IP network 100 configured according to the teachings of the present invention. Accordingly, a base station interface 140 may use two or more subnets 150,155 as entry points into the IP network 100. The IP network 100 is still configured for carrying data between the base station interface 140 and a public network interface 60. For example, both a first subnet 50 and a second subnet 55 are configured to transport data streams of traffic between a base station interface 40 and a public network interface 60. Thus, either, or both, of the first subnet 50 or the second subnet 55 may transport traffic between the base station interface 40 and the public network interface 60.

For example, the first subnet 50 could route traffic between the public network interface 60 and the base station interface 40. Likewise, the second subnet 55 could route traffic between the public network interface 60 and the base station interface 40. Accordingly, the subnets 50,55 function as alternative conduits, or alternative transport mechanisms, for carrying traffic through the IP network 100. Stated another way, the IP network 100 can be described as a transport mechanism for routing traffic between a public network 70 and a wireless network 7 through a plurality of subnets 50,55.

Furthermore, though two subnets 50,55 are shown in figure 2, an IP network according to the teachings of the present invention may incorporate any

number of subnets. Thus, the use of multiple subnets according to the present invention provides the ability to direct traffic between a base station interface and a public network interface across multiple subnets.

Figure 2 also shows an operations and maintenance (O&M) interface 154 which monitors the IP network 5 performance by gathering traffic data and statistics on the IP network 5 and the subnets 50,55. The O&M interface 154 sends the traffic data and/or statistics to a system control interface 152 which, in part, managing the flow of traffic across the IP network 5 based on the current traffic conditions and statistics. The O&M interface 154 may also establish new connections, or place additional features on the IP network 5.

Accordingly, the O&M interface 154 monitors the performance of the IP network 5 as a whole, and monitors the performance of the first subnet 50 and the second subnet 55 as well. In addition, the O&M interface 154 is adapted to process traffic statistic data to drive meaningful indicators relative to current traffic conditions.

The system control interface 152, comprises algorithms, such as those discussed below, for managing the flow of traffic across the IP network 5 based on traffic data, statistics, and other information provided by the O&M interface 154, the base station interface 40, the public network interface 60, or other data gathering devices. In addition, the system control interface 152 may control call placement and routing in the IP network 5. In one embodiment, the system control interface 152 and the O&M interface 154 are combined into a single node of the IP network 5.

The system of figure 2 routing calls through multiple subnets based on current network congestion levels. While the following descriptions of methods for practicing the present invention are directed particularly to a phone call, those of ordinary skill in the art will recognize that the present invention maybe configured for use with any other data types according to various embodiments. Furthermore, while the steps in Figures 3,4,5 and 6

are shown in a flow process format with sequential numbering, it will be apparent to those of ordinary skill in the art that several steps may be performed simultaneously and that some steps do not necessarily have to follow one after the other. When process ordering is essential practicing the invention, such a condition will be noted herein.

Figure 3 illustrates an algorithm for determining whether an IP network is carrying voice traffic. The process begins when a call request, step 210, is made by the mobile station 10. For example, the mobile station 10 could notify the IP network 100, using standard IP signaling sequences, that it wishes to place a new call. Then, in a voice traffic query 215, the system control interface will detect whether there is voice traffic present on the IP network 100. If there is no voice traffic on the IP network 100, then the algorithm proceeds to a first network processing algorithm, in step 217. If voice traffic is detected on the IP network 100, then the algorithm advances to a second network processing algorithm, in step 219.

Figure 4 is a process flow diagram of a method of placing a new call on the IP network 100 when is not presently carrying voice traffic (the first network processing algorithm). The process begins after a call request at step 210 is received and it is determined, in step 215, that the IP network 100 is not currently carrying voice traffic. Note, that other data formats could be traversing the network 100 at this point. An"arbitrary"choice of a subnet to carry the voice traffic might result in placing the voice traffic on a congested network. Accordingly, it is desired that the IP network 100 select a preferred subnet for carrying the new call.

Next, the method proceeds to gather traffic data in step 220. In step 220 the traffic data is gathered by data gathering devices such as an O&M interface 154 and/or the base station interface 40 at other nodes of the IP network 100. Accordingly, in step 220 the data gathering devices observe network traffic to obtain traffic data such as network congestion levels, call

failures, bit error rates, lost packet rate and other indicators of network congestion, for example. Step 220 may be implemented as an ongoing process in some IP networks, and may be simultaneously occurring with step 210 when a call request is made.

After traffic data is gathered, or statistics generated, the traffic data and/or statistics should be transferred, as shown by step 230. Accordingly, the data gathering devices send the traffic data to the system control interface 152 for analysis. In the event the O&M interface 152 and the system control interface 152 are implemented as a single interface, step 230 represents the transfer of traffic data (or statistics) from the traffic monitoring portion of the interface to the system control portion of the interface. Subsequently, the traffic data/statistics must be analyzed in steps 240 by the system control interface 152.

The system control interface 152 maintains the algorithms or tables for selecting a subnet based upon predetermined criteria, and may incorporate such concepts as statistical analysis, or statistical decision theory, for example. Then, in a subnet selection, step 250, the system control interface 152 selects a preferred subnet based on the predetermined criteria.

Following the selection of a preferred subnet in step 250, the system control interface 152 notifies the base station interface 40 which subnet is the preferred subnet in a subnet announcement step 260. In response to step 260, the base station interface 40 conducts packet preparation, step 270. In step 270 the base station interface 40 produces an appropriate frame header for properly routing a packet from the base station interface 40 to the public network interface 60 through the preferred subnet. Next, the base station interface 40, places the prepared packet onto the wireless telecommunications network 100 at the preferred subnet in a packet launching step 280.

Figure 5 is a block flow diagram of a method of placing a new call on an IP network 100 which is currently carrying voice data (the second network processing algorithm). The situation of an IP network already carrying voice traffic is differentiated from the situation of an IP network not carrying voice traffic because there is an overhead associated with sending any voice traffic over an IP network, regardless of the number of calls made. Accordingly, the present invention may save processing time by taking into account the transmission overhead associated with placing a voice transmission on the network 100 not currently carrying a voice transmission.

The method of figure 5, begins after a call request is received in step 210 and it is determined that the IP network 100 is already carrying voice traffic in the voice traffic query, step 215. Referring to figure 5, the data gathering devices gather traffic data in step 220 and transfers that traffic data in step 230 to the system control interface 152. In a subnet congested query step 232, the system control interface 152 checks the traffic data to determine whether the subnet that is carrying voice traffic is congested. If the subnet which is carrying voice traffic can handle addition voice traffic without a loss in quality, then, the subnet currently carrying voice traffic is used to carry the new phone call in a used subnet step 234. However, if the subnet that is carrying voice traffic is congested or appears to be in danger of becoming congested, then the method proceeds to analyze the traffic data in step 240.

After analyzing the traffic data in step 240, a preferred subnet is selected in step 250. Following the selection of a preferred subnet, the system control interface 152 informs the base station interface 40 of the selected subnet in step 260. The base station 40 next prepares the package which will carry the voice traffic by giving the packet (s) of voice traffic appropriate frame headers in step 270. Following the preparation of the packets in step 270, the base station interface 40 places the packet on the mobile telecommunications network 100 and packet launching step 280.

Figure 6 is a block flow diagram of a method of load leveling and reducing traffic congestion levels in the network according to one embodiment of the invention. The method of figure 6 can be adapted for maintaining high quality on the subnets of the IP network 100, rather than for introducing a new phone call onto the IP network 100.

Monitoring loop 200 provides for the periodic monitoring of traffic data as well as for detection of congestion states on the subnets 50,55 of the IP network 100. In a congestion query step 202 the system control interface 152 checks to see whether a congestion state exist within the IP network 100.

The congestion query step 202 could have the system control interface 152 monitor traffic data to detect indications of congestion, or may comprise the system control interface 152 receiving a"congestion state detected"signal generated by another system device.

If a congestion state is detected, the method immediately proceeds to a gather data step 220. However, if congestion is not detected in step 202, the method proceeds to check whether a predetermined amount of time has passed in a time query step 204. The predetermined amount of time is typically chosen by a system operator based on a number of factors such as the overhead involved with gathering and processing data, the number of subnets in the IP network 100, and a variety of other factors relevant to efficient allocation system resources. If the predetermined amount of time has not yet passed, the method returns to step 202. Otherwise, if the predetermined amount of time has elapsed, the method proceeds to gather data in step 220.

As in the steps of gathering data when a new call is placed on an IP network, in step 220 the data is gathered by data gathering devices such as an O&M interface 154 and/or the base station interface 40, for example.

Accordingly, in step 220 the data gathering devices observe network traffic to

obtain data such as network congestion levels, call failures, bit error rates, lost packet rate and other indicators of at work congestion.

After data is gathered, or statistics generated, they are transferred in step 230. Accordingly, the data gathering devices send the traffic data to the system control interface 152 for analysis. Subsequently, the traffic data/statistics are analyzed in steps 240 by the system control interface 152.

Then, in the subnet selection step 250, the system control interface 152 selects a preferred subnet based on the predetermined criteria.

In step 250, the system control interface 152 may reassign existing voice traffic, video traffic, or data traffic to any of the mobile telecommunication system subnets. For example, the subnet selection step 250 may place all voice traffic on a preferred subnet, and all additional traffic (meaning data, or video) on other subnets. Likewise, the subnet selection, step 250, may allocate traffic to subnet according to other factors, such as traffic data transfer speed, detected transfer rates at a destination public network, or other factors.

Following the allocation of traffic to subnets in step 250, the method proceeds to implement the selected network configuration by first announcing the subnet selection to the base station interface 40 in step 260. In response to step 260, the base station interface 40 conducts the packet preparation step 270 by producing appropriate frame headers for properly routing a packet (s) from the base station interface 40 to the public network interface 60 through the preferred subnet. Next, the base station interface 40, places the prepared packet onto the wireless telecommunications network 100 at the preferred subnet in a packet launching step 280.

While the invention has been described in conjunction with preferred embodiments, it should be understood that modifications will become apparent to those of ordinary skill in the art and that such modifications are therein to be included within the scope of the invention and the following claims.