System and method for achieving continuous connectivity to an access point or gateway in a wireless network following and on-demand routing protocol

Description of equivalent WO 2004064303 A2

0001Patent Application

0002for

SYSTEM AND METHOD FOR ACHIEVING CONTINUOUS

CONNECTIVITY TO AN ACCESS POINT OR GATEWAY L A WIRELESS

NETWORK FOLLOWING AN ON-DEMAND ROUTING PROTOCOL,

AND TO PERFORM SMOOTH HANDOFF OF MOBILE TERMINALS

BETWEEN FIXED TERMINALS LN THE NETWORK

0008by

0009Avinash Joshi

BACKGROUND OF THE INVENTION

0011Field of the Invention:

0012[OOOl] The present invention relates to a system and method for enabling an ad-hoc communication network to maintain connectivity within mobile nodes and fixed nodes in the network in an effective and efficient manner with minimal overhead. More particularly, the present invention relates to a system and method for enabling an ad-hoc communication network to maintain connectivity between intelligent access points of the network and other nodes (e.g., wireless routers and subscriber devices which can be fixed or mobile) in the network while using an on-demand routing protocol. The present invention also relates to a system and method for achieving smooth handoffs of mobile nodes between fixed nodes in an ad-hoc communication network using an improved distance vector routing algorithm and unicast messages. Furthermore, the present invention relates to a system and method for modifying the Ad Hoc On-Demand Distance Vector Routing (AODV) protocol to facilitate smooth handoff of subscriber devices in an ad-hoc communication network while also eliminating unidirectional links between nodes in the network. The present invention claims benefit under 35 U.S.C. § 119(e) from U.S. Provisional Patent Application Serial Nos. 60/439,448, 60/439,449 and 60/439,455 of Avinash Joshi, each filed on January 13, 2003, and from U.S. Provisional Patent Application Serial No. 60/476,237 of Avinash Joshi, filed on June 6, 2003, the entire contents of each being incorporated herein by reference. Description of the Related Art:

0013[0002] Wireless communication networks, such as mobile wireless telephone networks, have become increasingly prevalent over the past decade. These wireless communications networks are commonly referred to as "cellular networks", because the network infrastructure is arranged to divide the service area into a plurality of regions called "cells". A terrestrial cellular network includes a plurality of interconnected base stations, or base nodes, that are distributed geographically at designated locations throughout the service area. Each base node includes one or more transceivers that are capable of transmitting and receiving electromagnetic signals, such as radio frequency (RF) communications signals, to and from mobile user nodes, such as wireless telephones, located within the coverage area. The communications signals include, for example, voice data that has been modulated according to a desired modulation technique and transmitted as data packets. As can be appreciated by one skilled in the art, network nodes transmit and receive data packet communications in a multiplexed format, such as time-division multiple access (TDMA) format, code-division multiple access (CDMA) format, or frequency-division multiple access (FDMA) format, which enables a single transceiver at the base node to communicate simultaneously with several mobile nodes in its coverage area.

0014[0003] In recent years, a type of mobile communications network known as an "ad-hoc" network has been developed. In this type of network, each mobile node is capable of operating as a base station or router for the other mobile nodes, thus eliminating the need for a fixed infrastructure of base stations. Details of an ad-hoc network are set forth in U.S. Patent No. 5,943,322 to Mayor, the entire content of which is incorporated herein by reference.

0015[0004] More sophisticated ad-hoc networks are also being developed which, in addition to enabling mobile nodes to communicate with each other as in a conventional ad-hoc network, further enable the mobile nodes to access a fixed network and thus communicate with other mobile nodes, such as those on the public switched telephone network (PSTN), and on other networks such as the Internet. Details of these advanced types of ad-hoc networks are described in U.S. Patent Application Serial No. 09/897,790 entitled "Ad Hoc Peer-to-Peer Mobile Radio Access System Interfaced to the PSTN and Cellular Networks", filed on June 29, 2001, in U.S. Patent Application Serial No. 09/815,157 entitled "Time Division Protocol for an Ad-Hoc, Peer-to-Peer Radio Network Having Coordinating Channel Access to Shared Parallel Data Channels with Separate Reservation Channel", filed on March 22, 2001, and in U.S. Patent Application Serial No. 09/815,164 entitled "Prioritized-Routing for an Ad-Hoc, Peer-to-Peer, Mobile Radio Access System", filed on March 22, 2001, the entire content of each being incorporated herein by reference.

0016[0005] As can be appreciated by one skilled in the art, since certain nodes of the ad-hoc network are mobile, it is necessary for the network to maintain connectivity with those nodes. Accordingly, needs exist for improved techniques to enable an ad-hoc network to maintain connectivity with the mobile nodes in the network in an effective and efficient manner with minimal overhead. Similarly, most of the traffic flows through the Access Point (AP) in such a network and hence, there exist a need for all nodes to maintain routes with the Access Point (AP) all the time in an effective and efficient manner with minimal overhead.

0017SUMMARY OF THE INVENTION [0006] An object of the present invention is to provide a system and method for enabling an ad-hoc communication network to maintain connectivity within the mobile nodes and fixed nodes in the network in an effective and efficient manner with minimal overhead. [0007] Another object of the present invention relates to a system and method for achieving smooth handoffs of mobile nodes between fixed nodes in an ad-hoc communication network.

0018[0008] These and other objects are substantially achieved by providing a system and method for enabling an ad-hoc communication network to maintain connectivity between intelligent access points of the network and other nodes (mobile nodes and fixed nodes) in the network while using an on-demand routing protocol.

0019[0009] These and other objects are further substantially achieved by providing a system and method, for use in an ad-hoc communication network, which enables the network to perform smooth handoffs of mobile nodes between fixed nodes in an ad-hoc communication network using an improved distance vector routing algorithm and unicast messages, thus avoiding an increase in routing advertisement frequency while keeping overhead at a minimum.

0020[0010] This and other objects are still further substantially achieved by providing a system and method for modifying the Ad Hoc On-Demand Distance Vector Routing (AODV) protocol to facilitate smooth handoff of subscriber devices in an ad-hoc communication network while also eliminating unidirectional links between nodes in the network.

0021BRIEF DESCRIPTION OF THE DRAWINGS [0011] These and other objects, advantages and novel features of the invention will be more readily appreciated from the following detailed description when read in conjunction with the accompanying drawings, in which:

0022[0012] Figure 1 is a block diagram of an example ad-hoc packet switched wireless communications network including a plurality of nodes in accordance with an embodiment of the present invention;

0023[0013] Figure 2 is a block diagram illustrating an example of a mobile node employed in the network shown in Figure 1;

0024[0014] Figure 3 is a conceptual block diagram of an example of the relationship between wireless routers and access points of the network shown in Figure 1;

0025[0015] Figure 4 is a conceptual block diagram of an example of the relationship between a moving subscriber device with respect to stationary wireless routers of the network shown in Figure 1 ;

0026[0016] Figure 5 is a conceptual block diagram of an example of the relationship between a moving subscriber device with respect to stationary wireless routers and access points of the network shown in Figure 1; and

0027[0017] Figure 6 is another conceptual block diagram of an example of the relationship between a subscriber device, wireless routers and an access point of the network shown in Figure 1. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0018] Figure 1 is a block diagram illustrating an example of an ad-hoc packet-switched wireless communications network 100 employing an embodiment of the present invention. Specifically, the network 100 includes a plurality of mobile wireless user terminals 102-1 through 102-n (referred to generally as nodes 102, mobile nodes 102 or subscriber devices (SD)), and can, but is not required to, include a fixed network 104 having a plurality of intelligent access points 106-1, 106-2, ...106-n (referred to generally as nodes 106, access points 106 or IAPs), for providing nodes 102 with access to the fixed network 104. The fixed network 104 can include, for example, a core local access network (LAN), and a plurality of servers and gateway routers to provide network nodes with access to other networks, such as other ad-hoc networks, the public switched telephone network (PSTN) and the Internet. The network 100 further can include a plurality of fixed routers 107-1 through 107-n (referred to generally as nodes 107, fixed routers 107 or wireless routers (WR)) for routing data packets between other nodes 102, 106 or 107. It is noted that for purposes of this discussion, the nodes discussed above can be collectively referred to as "nodes 102, 106 and 107", or simply "nodes".

0028[0019] As can be appreciated by one skilled in the art, the nodes 102, 106 and 107 are capable of communicating with each other directly, or via one or more other nodes 102, 106 or 107 operating as a router or routers for packets being sent between nodes, as described in U.S. Patent No. 5,943,322 to Mayor, and in U.S. Patent Application Serial Nos. 09/897,790, 09/815,157 and 09/815,164, referenced above.

0029[0020] As shown in Figure 2, each node 102, 106 and 107 includes a transceiver 108 which is coupled to an antenna 110 and is capable of receiving and transmitting signals, such as packetized signals, to and from the node 102, 106 or 107, under the control of a controller 112. The packetized data signals can include, for example, voice, data or multimedia information, and packetized control signals, including node update information. [0021] Each node 102, 106 and 107 further includes a memory 114, such as a random access memory (RAM), that is capable of storing, among other things, routing information pertaining to itself and other nodes in the network 100. The nodes 102, 106 and 107 take part in routing protocol which can be on demand or proactive and hence can send routing messages which can be Route Request, Route Reply, Route Error or Routing Advertisement whenever there is a change in the network topology or on a periodic basis. [0022] As further shown in Figure 2, certain nodes, especially mobile nodes 102, can include a host 116 which may consist of any number of devices, such as a notebook computer terminal, mobile telephone unit, mobile data unit, or any other suitable device. Each node 102, 106 and 107 also includes the appropriate hardware and software to perform Internet Protocol (IP) and Address Resolution Protocol (ARP), the purposes of which can be readily appreciated by one skilled in the art. The appropriate hardware and software to perform transmission control protocol (TCP) and user datagram protocol (UDP) may also be included.

0030[0023] As can be appreciated from the above, in the network 100, an LAP 106 is the point of attachment of the wireless part and the wired Internet. If a Subscriber Device (SD) 102 is not in direct communication range with an LAP 106, the DS 102 depends upon other devices to reach the IAP 106. These devices can be other SDs 102 or Wireless Routers (WR) 107 which are deployed specifically to provide coverage to these SDs.

0031[0024] In an ad-hoc network such as network 100, all nodes need to maintain continuous connectivity with the Access Point (or LAP) as most of the traffic is to and from an IAP. This is true for both the field deployment and deployment in an office environment. Continuous connectivity is also needed to tunnel the dynamic host configuration protocol (DHCP) and address resolution protocol (ARP) IP broadcast to the LAP. If an on-demand protocol, such as the Ad Hoc On-Demand Distance Vector Routing protocol (AODV), is used in an ad-hoc network, no routes are maintained proactively. Accordingly, the embodiments of the present invention described herein provide routes to IAPs that can be maintained at all times with minimum overhead. This technique can also be used as an LAP association protocol. The technique has been named Proactive LAP Locator (PIL) protocol.

0032[0025] As can be appreciated by one skilled in the art, an on-demand routing protocol (for example AODV) creates routes only when desired by the source node. AODV is described in a publication by Charles E. Perkins, Elizabeth M. Belding-Royer, and Samir Das entitled "Ad Hoc On Demand Distance Vector (AODV) Routing", RFC 3561, July 2003, the entire contents of which are incorporated herein by reference.

0033[0026] Hence, when a node requires a route to a destination, the node initiates a route discovery process within the network 100. The commonly employed approach, referred to as an Expanding Ring Search, can increase the average latency of route discovery, as multiple discovery attempts and time-outs may be needed before a route to the target node is found, the high latency will require the source node to buffer the packets which may be difficult for memory constrained nodes in such a kind of network resulting in packet loss. Furthermore, this elongated route discovery process also increases the overhead as each discovery can lead to a network wide flood. Since most traffic in this kind of network flows between LAP and other nodes, such as WRs and SDs, these floods can be avoided if the nodes proactively maintain routes to the LAP. This will also avoid the latency and buffering of packets involved in finding route to an LAP. AODV also assumes bi-directional links among nodes, which can lead to incorrect routes. The technique according to the embodiments of the invention described herein avoids unidirectional links while finding routes from nodes to the LAP, and vice-versa.

0034[0027] An example of the process will now be described with reference to Figure 3. hi this example, it is assumed that all the nodes follow AODV routing protocol with the modifications in accordance with the embodiments of the present invention described herein. All the nodes will periodically broadcast a packet that is referred to as IAP Advertisements ( A). Ln place of an LA packet, the nodes can also use a regular "Hello Message" which is generally sent by all nodes in such network to maintain connectivity as described, for example, in a U.S. Provisional Patent Application entitled "System and Method to Maximize Channel Utilization in a Multi-Channel Wireless Communication Network", Serial No. 60/475,882, filed on June 5, 2003, the entire contents of which being incorporated herein by reference. [0028] The contents of the IA packet can be any combination of the following fields:

0035• Type of node : This will let other nodes know about the type of the device which can help them decide whether or not this node should be used to route packets. This field can also help in deciding the routing metrics as described in U.S. Provisional Patent Application Serial No. 60/476,237 referenced above.

0036• Number of hops from the associated IAP.

0037• Address of the node (IP or MAC address or both): This decision is based on whether the network uses layer 2 routing or layer 3 routing or a combination of both. • Address of the associated LAP (IP or MAC address or both): This decision is based on whether the network uses layer 2 routing or layer 3 routing or a combination of both.

0038• Routing Metrics to the associated IAP: This field helps in deciding one route versus another as described, for example, in U.S. Provisional Patent Application Serial No. 60/476,237 referenced above, and in a U.S. Patent Application of Eric A. Whitehill et al. entitled "EMBEDDED ROUTING ALGORITHMS UNDER THE INTER-NET PROTOCOL ROUTING LAYER OF A SOFTWARE ARCHITECTURE PROTOCOL STACK", Serial No. 10/157,979, filed on May 31, 2002, the entire contents of which are incorporated herein by reference.

0039• Other Metrics (for example some metrics representing the load on the LAP like number of active users associated with the LAP or total bandwidth used by the users): this field can be used to do load balancing across multiple LAPs and achieve quality of service (QoS) goals across multiple LAPs.

0040• QoS metrics: Used for QoS Routing.

0041• Address of the node that is being used as next hop towards the LAP: This field can be used to perform a "Split Horizon" technique as done by classical distance vector protocols in the Internet as described below.

0042• Broadcast LD: a broadcast ID similar to that used in the AODV route request (RREQ) process, which is helpful in detecting duplicate packets and dropping those duplicates. This broadcast LD may not be required if some sequence number is already a part of MAC header to discard duplicate packets.

0043• Power Level: This message can either be sent at some fixed power that is known throughout the network, or power used should be indicated in this field of the packet. This will help the node receiving the packet to know the path loss between the transmitter and itself.

0044• TTL: The packet can also have a (time-to-live) TTL value set to NETWORK_DIAMETER, which depends upon the size of the network 100 and the maximum number of hops possible between an IAP and a node associated with that LAP. The TTL value can then be decremented by each protocol interface layer (PIL) daemon as the packet propagates throughout the network 100, which can control the maximum number of hops possible between an IAP and any other node in the network. • A node can also send similar information about some other IAP/IAPs with which it is not associated. The information can include all the metrics mentioned above. [0029] Referring to Figure 3, the process is started when an LAP broadcasts one of these packets (step 1), and the nodes in the network 100 which are one hop away from the LAP receives it. On receiving such a message, the node stores the relevant information from the message in a table and compares that information against other entries if it has received the same message from other neighboring nodes. It can also simply discard the message if the next hop field in such a packet has its own address. This is done to avoid loops and is similar to the well-known "Split Horizon" method used in conjunction with Distance Vector Routing Protocol. As can be appreciated by one skilled in the art, Split Horizon is a well known method in wired networks to solve the "count-to-infinity" problem in distance vector protocol. Although the algorithm is not perfect and the problem of "count-to-infmity" can still exist if the number of nodes involved are more than 3, it is still useful. The algorithm selectively excludes a destination from an advertisement to a neighbor, if the next hop to that destination is that neighbor. Since the LA packet is a broadcast packet and sent in place of a unicast packet to an individual neighbor, the split horizon method cannot be used in its original form. Instead, the next hop used to reach the destination (an LAP in the present example) can be advertised in the packet. Hence, when a neighbor receiving the packet sees its address in the next hop field, it ignores the advertisement to avoid the "count-to-infinity" problem. Thus, the split horizon method is used on the receiver side in wireless network as opposed to being used on the transmitter side as in a wired network.

0045[0030] These nodes will now make a routing decision based on number of factors, including but limiting to, the number of hops, routing metrics, load balancing metrics, QoS metrics, and so on. If a node decides to use the sender of this message as a next hop towards the LAP (the LAP itself in the current example), then it sends a unicast Route Request (RREQ) for the LAP address to the node (the LAP in this case) which forwarded the IA packet (step 2). Unicasting the RREQ will confirm that the link is not a unidirectional link, but rather, a bidirectional link. If such a Route Request fails, the node can blacklist the sender for some time and wait for IA messages from other nodes so that the node can hop through those other nodes to reach LAP. The nodes that are successful in the route discovery process receive a Reply (RREP) back from the LAP (step 3), and the LAP also creates a reverse route back to the node following the normal AODV routing protocol. This reverse route can be used to send an LAP association/update message.

0046[0031] After obtaining the route, the nodes rebroadcast the IA message (step 4) after updating the relevant fields like increasing the hop count, decrementing the TTL, updating the different metrics etc. Nodes receiving this IA will repeat the procedure by unicasting a RREQ (step 5) to the forwarding node and receive a unicast Reply (RREP) from the forwarding node (step 6). It is noted that the G bit in this unicast RREQ will be set so that a Gratuitous RREP is also sent (step 7) to the destination node (in this case, the LAP), so that the LAP also learns of a route to the nodes. In this manner, knowledge of the LAP (which is the default route in network 100) is proactively flooded over the network 100. As indicated in step 8, the nodes receiving the rebroadcast IA then themselves re-broadcast the IA (step 8). [0032] In accordance with the AODV protocol, after a node receives a RREQ and responds with a routing reply RREP, the node discards the RREQ. If intermediate nodes, such as other SDs or WRs, reply to every transmission of a given RREQ, the destination (e.g., IAP) does not receive any copies of the RREQ. In this situation, the destination does not learn of a route to the originating node. In an ad-hoc network, if WRs always reply to the RREQs, the LAP will never learn about any route to the SD. Currently, the AODV draft has a provision to let the destination know about this route. Specifically, the AODV draft states that in order that the destination learn of routes to the originating node, the originating node SHOULD set the "gratuitous RREP" ('G') flag in the RREQ. If, in response to a RREQ with a 'G' flag set, an intermediate node returns a RREP, it must also unicast a gratuitous RREP to the destination node. Charles E. Perkins, Elizabeth M. Belding-Royer, and Samir Das. "Ad Hoc On Demand Distance Vector (AODV) Routing", referenced above. This is the reason for enabling the G bit in the RREQ packet. However, in place of the G bit, a D bit (Destination only flag) can also be set so that noone other than destination (LAP in this case) replies to the message.

0047[0033] A summary of the Proactive LAP Locator Algorithms are as follows. [0034] A Proactive LAP Locator Daemon of an LAP gets the IP or MAC address of the LAP, and builds an LAP Advertisement packet containing data pertaining to an UP/MAC address of the LAP. The associated LAP is set to 0 as it is itself an LAP, the next hop towards the LAP is also initialized to 0, the number of hops set to 0, and appropriate metrics are set. If the Network Broadcast ID is used it is initialized to 0, and if the power used for the message is not fixed, the quantized value of the power is placed in the power level fields, and the TTL is set to the NETWORK_DLAMETER value. The Proactive LAP Locator Daemon then broadcasts the LAP Advertisement packet on all interfaces. The Proactive LAP Locator Daemon then repeats LAP Advertisement every LAP_ADVERTISEMENT_INTERVAL seconds, which is a configurable parameter, while incrementing the Network Broadcast ID. [0035] A Proactive LAP Locator Daemon of the WRs and SDs listens for an LAP Advertisement packet containing data pertaining to an LP/MAC address of an LAP, a number of hops from the forwarding node, different metrics, and so on, as mentioned above. If the node decides to use the sender as the next hop towards the LAP, the Proactive IAP Locator Daemon issues a RREQ for LAP with a G bit or D bit set to the forwarding node. If an RREP is received, the Proactive LAP Locator Daemon sends an association/update message to the IAP, increments the Hops field, decrements the TTL, update other fields, and forwards to broadcast address on all interfaces. However, if no RREP is received, Proactive LAP Locator Daemon waits for another IA packet and may elect to blacklist the sender for some time. It should be also noted here that all nodes send the IA or Hello message on a periodic basis which can be configurable based on device type or other factors.

0048[0036] As can be appreciated by one skilled in the art, the technique according to the embodiments of the present invention described above avoids the high latency prone route discovery process for a common destination, such as an LAP, while minimizing overhead in the network 100. The technique can be used as an LAP association protocol, and load balancing can be achieved among different LAPs serving the network 100. [0037] It is also noted that unicasting techniques similar to those discussed above can be used to provide smooth handoff between mobile SD, such as those being used in a moving vehicle. Figure 4 shows a typical scenario where wireless routers WR1 and WR2 are deployed to provide coverage on a highway. The circles represent the range of these WRs, which means that any device, such as an SD, LAP or another WR will be able to communicate with this WR if it is inside this circle. As it can be seen in the figure, there is a small area where these two circles overlap. Ln this area, the SD is in position to communicate with both the wireless routers. It is advantageous for service providers to keep this area as small as possible as it reduces the number of WRs that needs to be deployed in a given area. [0038] When a mobile SD is moving at highway speed, it rapidly crosses the coverage area of the wireless routers WR1, and hence, passes by the overlap area in a very small period of time. During this small time period, the SD needs to change its routing table to reflect the fact that its best next hop to the infrastructure has changed from WR1 to WR2. A similar process is referred to as "handoff in cellular networks.

0049[0039] In an ad-hoc such as network 100 described above, most of the traffic flows between an LAP and the SD, so in this small period of time, the LAP should also be informed about this change of point of attachment of SD to the network 100. The embodiments of the invention described herein provide a system and method to achieve this and other objectives in a fast and efficient manner. Specifically, the embodiments provide a method which enables an SD to update its routing table to reflect the change in its point of attachment to the network 100 (that is, its affiliation with a specific LAP) in a fast and efficient way with minimum loss of packets (called "Smooth Handoff). An example of a smooth handoff technique is disclosed in U.S. patent application of Robin U. Roberts and Charles R. Barker, Jr. entitled "A System and Method for Performing Soft Handoff in a Wireless Data Network", Serial No.09/929,031, filed on August 15, 2001, the entire contents of which is incorporated herein by reference. The embodiments also provide a method for other devices (in particular, IAP 106) to know about this change in a fast and efficient manner. [0040] In a Distance Vector approach, devices such as SDs, IAPs and WRs learn about the changes in routes through periodic routing advertisements. Thus, it takes substantial time (depending upon the periodic interval between these advertisements) before an SD can know that it has moved away from a WR and is close to another WR. This elapsed time is even greater for an LAP which is typically several hops away from the place where routes have changed. This is illustrated in the following example which will be discussed with reference to Figure 5 in particular.

0050[0041] As shown in Figure 5, a mobile SD is moving at a high speed on a highway where coverage has been provided by wireless routers WR-3 through WR-6. In this example, the SD moves from left to right in Figure 5 as shown, meaning that it is moving from the coverage area of WR-3 to WR-4. It is assumed that the time at which the mobile (SD) comes in communication range of WR-4 is t, and that the periodic interval between two consecutive Routing Advertisements is T (which is same for SDs and WRs). It is also assumed that there is a three way handshake mechanism which confirms that SD has indeed moved into the coverage range of WR-4 and there exist a bi-directional link between them. This exchange can take up to ST periods. After this handshake, WR will start advertising the SD. The information will be propagated in following manner (in a worst case scenario)

0051Time (t + 3T): WR-4 is done confirming the presence of SD in its wireless range

0052Time (t + 4T): WR-4 sends this information in its RA, WR-1 receives this information.

0053Time t + 57): WR-1 sends this information in its RA, LAP now knows about this

0054[0042] Thus it takes (hops + 2) times periodic time of RA to inform the LAP about the change of route of SD, where hops = number of hops between SD and IAP [0043] If T is large it will take a long time for the LAP to know about the route change. Also, the time needed to perform the three way handshake can also be so large that the SD actually moves out of the range of the WR involved in the handshake before the handshake process has been completed. One way to deal with this issue is to have a lower T. However, having a lower T substantially increases the overhead of the network. The embodiments of the invention described herein thus provide a technique in which this handshake operation can be performed in a more efficient way so that the information can be propagated to the LAP without major overhead.

0055[0044] To perform this technique, all infrastructure devices, that is, all LAPs and WRs, are required to periodically broadcast a Neighbor Advertisement or Hello Message. This message can have similar fields as in the IA packet mentioned above. A three way handshake protocol can be used here to confirm the bidirectional link.

0056[0045] All SDs actively monitor this Neighbor Advertisement or Hello Message and measure the signal strength of the signal. An SD can also do the three way handshake to determine the received signal strength indicator (RSSI) and post detection signal quality (PDSQ) values on both ends. As soon as the SD determines from the signal strength and/or these values that it will loose connectivity with the old WR and should handoff, the SD unicasts an empty Routing Advertisement to the WR to which it wants to handoff. It can also decide to handoff based on the routing metrics advertised in the Routing Advertisement or calculated by the node. In the example shown in Figure 5, if the SD is moving from the coverage area of WR-3 to WR-4, the SD will send this unicast routing advertisement (RA) to WR-4. This RA will be empty, that is, there will be no other entries for other nodes but will have the header which will inform the receiver node about this SD. The WR (i.e., WR-4) receiving this unicast RA will update its routing table in normal way, but since this was a unicast RA, it will also unicast one RA to the next hop towards the IAP with which it is affiliated. In this example, upon receiving the RA from SD, WR-4 will consult its routing table to find the best next hop towards LAP, and will send a unicast RA to that node. This RA will just have one entry about the SD. In this example, the best next hop to LAP is WR-1, which will follow the same procedure to direct the unicast RA towards the IAP. Other message types can also be used in place of explicit RA, for example some geo packet can also be used to carry the information carried by routing advertisement. The information can sometimes be piggy backed to data packet as well.

0057[0046] As can be appreciated from the above, the SD can thus handoff in a very small time which does not depend upon the periodic RAs. Also, the LAP will become aware of the movement of SD in real time rather than waiting to receive a periodic RA. Since this critical route update no longer depends upon the broadcast RA, the periodic interval can be increased which will result in substantially lower overhead. Hence, the embodiment achieves fast route convergence and low overhead because the frequency of transmission of Routing Advertisements can be reduced.

0058[0047] Another smooth handoff technique, which is a modification to Ad Hoc On- Demand Distance Vector Routing (AODV), will now be described with regard to Figure 6. [0048] As can be appreciated by one skilled in the art, AODV is a well known on- demand routing protocol. Specifically, this type of routing protocol creates routes only when desired by the source node. When a node, such as an SD, WR or LAP, requires a route to a destination, such as another SD, WR or LAP, the node initiates a route discovery process within the network 100. This process is completed once a route is found or after all possible route permutations have been examined. Once a route has been established, the established route is maintained by some form of route maintenance procedure until either the destination becomes inaccessible along every path from the source or until the route is no longer desired. [0049] Although the on demand approach reduces the routing overhead, it adds latency in obtaining a route since the routes are not computed before they are actually needed. Because of latency involved in finding routes, packets are needed to buffer the packets at the source. If the node is memory constrained, packet losses can occur due to latency. Therefore, this approach generally does not facilitate smooth handoffs.

0059[0050] The following describes modifications to the standard AODV technique according to embodiments of the present invention, which facilitate smooth handoffs in an ad-hoc network. These modifications also help to eliminate unidirectional links in the network 100, because AODV assumes bidirectional links.

0060[0051] Figure 6 illustrates a portion of the network 100 in which Wireless Routers WR-1 and WR-2 are connected to LAP via wireless links. The LAP is connected to the core LAN as shown in Figure 1. An SD is also shown which moves at a highway speed. [0052] It can be assumed that most of the time, the SD will need a route to the LAP because the major applications performed by the SD will be Internet browsing, Voice Over Internet Protocol (VOIP) phone calls, and so on. By using the standard AODV routing protocol, much time is consumed before the route discovery process is completed. Furthermore, by the time the SD will receive a route reply, it is probable that the SD has moved to different position which may be outside of the range of a particular WR included in the route reply. The following technique according to an embodiment of the present invention avoids this drawback and helps in achieving smooth-handoffs of the SD from one WR or LAP to another.

0061[0053] In this example, it is assumed that all WRs and SDs are always associated with one LAP. LAPs and WRs periodically transmit a "Hello Message" (or Neighbor Advertisement) which can have fields similar to IA packet described above. The fixed nodes can maintain the routes towards the LAP by sending a periodic RREQ as mentioned above. These nodes can also maintain the route by just sending the RREQ the first time and then by not expiring the route towards the LAP after getting the Route Reply.

0062[0054] To provide smooth handoff, the SDs will send an RREQ for their associated LAP at a periodic rate. This periodic interval is variable depending upon various factors such as speed of the vehicle in which the SD is disposed (if the speed can be determined, for example, by noting the average change of geo location, the activity of the SD, the rate of change of signal quality with the neighboring WRs, and an average number of new neighbors added in the neighbor table). The Route Request packets are generated with either the G bit or D bit set. Accordingly, the destination (LAP) learns about the route without doing any Route Request. If a packet arrives at an LAP destined to some SD, the packet can be delivered because the LAP knows the routes to all of the SDs which are associated with that IAP.

0063[0055] It is also noted that the WRs should periodically broadcast neighbor advertisements in which they should include their IP address as well as that of the LAP with which they are associated. The SD should cache these advertisements and should periodically unicast the RREQ to these WRs. If the link quality measurements are continuously made through passive listening of request-to-send/clear-to-send (RTS/CTS) or these advertisements, the SD can intelligently chose, for example, one to three 1-3 of the WRs it is currently listening to (depending upon the number of back up routes desired) and unicast an RREQ to those WRs. As stated, the number of WRs need not be within the range of one to three, but rather, can be any suitable number based on the number of back up routes desired.

0064[0056] In place of sending periodic Route Request (RREQ) to several wireless routers, all the nodes in the network can simply use the Proactive LAP Locator protocol and achieve smooth handoff of mobile nodes. For this to work, the nodes should actively receive the IA/hello message/neighbor advertisement and should use such a combination of metrics that signify good route in terms of throughput as well the signal strength between the node and the next hop. For example, if the metrics involve the combination of hops, node types, data rate, signal strength, battery power, packet delivery ratio and so on, then as a mobile node moves away from a WR (i.e., WR2) and moves closer to another WR (i.e., WR1) as shown in the Figure 6, the cumulative routing metrics between the mobile node and the IAP through WR2 will increase while one through WR1 will decrease. As stated, it is assumed in this example that the mobile node is moving away from the coverage area of WR2 and is moving into the coverage area of WR1. Also, it is assumed in this example that lower routing metrics represent a better route than higher routing metrics. Accordingly, the mobile node will automatically switch the routes and send an RREQ to WR1 on determining that the routing metrics through WR1 are better than through WR2, to thus achieve smooth handoffs. [0057] As can be appreciated by one skilled in the art, this scheme has the advantage of eliminating unidirectional links. As discussed above, the AODV protocol assumes bidirectional links and hence, a network 100 operating in accordance with AODV creates a reverse route from a node that receives a RREQ to the node that made the RREQ as soon as the receiving node receives the RREQ. It should be noted that if this link is not bidirectional but rather, unidirectional, the AODV protocol can erroneously create incorrect routes. However, unicasting the RREQ message confirms that the link is bidirectional since the hand-shake of RTS-CTS is completed between the node that is sending the RREQ and the node that is to receive the RREQ before the sending node sends the RREQ. Furthermore, the bidirectional link is verified because the node sending the RREQ receives an acknowledgement (ACK) back for the unicasted RREQ from the node that receives the RREQ.

0065[0058] Accordingly, the technique described above provides for smooth handoffs of SDs and low latency in the network 100. Furthermore, each LAP in the network 100 knows about the routes to all SDs associated with itself all the time, and these routes between the LAP and its associated SDs are updated in real time, which can help in reducing the route setup time for phone calls originating from outside network, for example. Furthermore, overhead in the network 100 is minimized because the periodic RREQ can be piggy backed with the data packets which are already being sent from SDs, WRs and so on.

0066[0059] In addition, although all of the techniques described above are used with AODV in the examples given, these techniques can be used with other routing protocols and, in particular, on-demand type routing protocols such as dynamic source routing (DSR) or any other suitable protocol.

0067[0060] Although only a few exemplary embodiments of the present invention have been described in detail above, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the novel teachings and advantages of this invention. Accordingly, all such modifications are intended to be included within the scope of this invention.