KR20090080355A - Method for a load balancing scheme with route cache management - Google Patents

Abstract

A load balancing method using route cache information is provided to enhance efficiency in setting a routing path by preventing traffic from concentrating on a node. A first node transmits an RREQ(Route Request) packet for setting a routing path to a second mode. The second node receives the RREQ packet and determines whether the second node is a destination node or not(S211). When the second node is not the destination node, it is determined whether routing information to the destination node is found in a route cache(S231). When the routing information is found, load balancing is performed by compares a buffering value, which is the number of packets buffered in a queue, with a threshold buffering value and determining to transmit an RPEP packet to the first node based on a comparison result(S251).

Description

Load balancing method using route cache information {Method for a load balancing scheme with route cache management}

The present invention relates to a load balancing method using path cache information, and to a load balancing method using path cache information that prevents traffic concentration of a specific node and provides load balancing using information stored in the path cache.

A sensor network is a wired / wireless network composed of many ultra-lightweight, low-power sensors for ubiquitous computing, and basically consists of a sensor node and a sink node. The sensor node is a system composed of a sensor, a processor, and a communication element that transmits or transmits information sensed from an environment and a physical system or a specific event related to a sensor based on wired or wireless communication technology. Each sensor node (mobile or source node) floats a RREQ (Route Request) packet to send the collected data to the destination node. The RREQ packet is a control message type used by the source node to find the destination node. The node generating the RREQ packet can be an intermediate node that initiates a path recovery operation for path recovery when a path break occurs with the source node that originally requested data delivery. These nodes will then float the RREQ packet into the Ad-Hoc network. In addition, nodes that receive an RREQ packet have a unicast (source node) to the node (source node) that originally sent the RREP (Route Reply) message if the route is present in the route (routing) table they have or if they are the destination node. unicast). Also, if the node receiving the RREQ packet is not the destination node and does not know the path to the destination node, store the reverse route to the source node in the route table and broadcast the RREQ packet to the neighbor node. do. When the path is established through the above process, the number of hops from each node to the destination node is determined, and then data transmission is performed through the configured routing path.

1 illustrates a path setting method between a conventional source node and a destination node.

As shown in FIG. 1, the source (sending) node N10 broadcasts an RREQ packet to the neighbor node N20 to search for a path to the destination (receive) node N30. The intermediate node N20 receiving the RREQ packet searches for route information from its route table to the destination node N30. If there is route information, the information is transmitted to the source node N10. If there is no route information, the RREQ packet is rebroadcasted to the neighbor nodes N21, N24, and N26 so that the RREQ packet can be delivered to the node or destination node that knows the destination node N30. Reverse path information (N20-> N10) to node N10 is stored. In this manner, the intermediate nodes N21 to N28 forward the RREQ packet to the destination node N30. When the RREQ packet arrives at the destination node N30, the destination node N30 takes a path (N10-> N20-> N24-> N25-> N30) with the smallest hop count among the various paths to which the RREQ packet has been forwarded. It generates a RREP packet so that the source node (N10) knows the corresponding path, and unicasts to the source node (N10) by reversing the path through which the RREQ packet is delivered.

In FIG. 1 described above, when the new source node N11 searches for a path for delivering data to the destination node N30, the source node N11 transmits an RREQ packet to the neighbor node N20 in the same manner as described above. Broadcast to. Since the intermediate node N20 that has received the RREQ packet has information about the destination node N30, the intermediate node N20 does not propagate the RREQ packet anymore, and thus delivers route information to the new source node N11 to the destination node N30. . As a result, the two source nodes N10 and N11 share the same path and forward the packet. In addition, even if the RREQ packet is delivered to the destination node N30 because the intermediate nodes do not know the path, the existing path that considers only the minimum hop counts as the path metric in setting the path is also the same path (N11−). > N20-> N24-> N25-> N30). As a result, traffic is concentrated only on intermediate nodes N20, N24, and N25 on the shared path.

Meanwhile, EAODV proposes an optional route caching method for traffic route distribution. That is, the derivation of the RREQ packet is induced by using the source / destination address of the cache to distribute the traffic path. However, the bypass of the RREQ packet unnecessarily lengthens the path, thereby causing a problem of transmission delay.

In addition, “Workload Based Load Balancing” intentionally defines the amount of traffic concentrated on a specific node as “Workload” and selects a specific threshold to drop the RREQ packet when it exceeds the threshold, intentionally in the path discovery process. Suppress the participation of the node. This reduces the floating number of RREQ packets and achieves load balancing by routing through other nodes. However, a drop of the RREQ packet without considering the situation of other nodes in the neighbor may lead to a path discovery failure, which may result in network separation.

Therefore, there is a need to efficiently solve the traffic concentration of a specific node and to reduce the packet transmission delay and buffer overflow for data communication through load balancing.

An object of the present invention is to provide a load balancing method using path cache information, to prevent traffic concentration of a specific node, and to provide load balancing using path cache information.

In addition, when the RREQ packet includes a QoS flag for providing priority information, the present invention omits the step of comparing the buffering value and the buffer threshold with reference to the priority, and includes information on the destination node. The purpose is to improve the efficiency of routing path by sending the RREP packet to the source node.

In order to achieve the above object, the load balancing method using the route cache information according to an embodiment of the present invention comprises the steps of (a) the first node transmits a Route Request (RREQ) packet for routing path establishment to the second node and (B) determining whether the second node receives the RREQ packet and whether the second node is the destination node, and if the second node is not the destination node, route information to the destination node in the route cache. In step (c) of determining whether there is a path and the path information to the destination node, the second node compares the buffering value, which is the number of packets buffered in the queue, with a predetermined buffer threshold value, and then the RREP ( And (d) performing load balancing by determining whether to transmit the packet to the first node.

According to the load balancing method using the path cache information of the present invention has the following advantages.

First, by comparing the buffering value, which is the number of packets buffered in the queue of a given node, with the buffer threshold value, if the buffering value is larger than the buffer threshold value, the transmission of the RREP packet to the source node is suppressed and the RREQ packet to the neighbor node. By retransmitting, there is an advantage that can prevent traffic concentration of a specific node.

Second, if the RREQ packet contains a QoS flag that provides information about the priority, the node containing the information about the destination node, omitting the step of comparing the buffering value to the buffer threshold with reference to the priority, By sending RREP packet to source node, routing routing establishment efficiency can be improved.

Third, in an ad hoc network with high mobility and large change in radio link, if the buffering value is larger than the buffer threshold, the node may acquire the latest path information by retransmitting the RREQ packet to the neighbor node. .

Fourth, the packet set the QoS flag is not dropped compared to other packets, it is also possible to achieve a fast routing path setting using the priority information included in the QoS flag.

Advantages and features of the present invention and methods for achieving them will be apparent with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but can be implemented in various different forms, and only the embodiments make the disclosure of the present invention complete, and the general knowledge in the art to which the present invention belongs. It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims. Like reference numerals refer to like elements throughout.

On-demand protocols such as AODV, DSR, etc. have been proposed to efficiently route nodes on a mobile Ad-Hoc network. To do this, the source node broadcasts an RREQ packet and receives an RREP packet from the destination node. At this time, the transmission of the RREQ packet is the best method to perform when the node does not know the path information in the dynamic network topology. In addition, the path cache is used to reduce network bandwidth waste and reduce search delay. At this time, the intermediate node receiving the RREQ packet refers to the information of its route cache, and if there is information that matches the destination address in the route cache, the intermediate node responds to the source node with the RREP packet.

In addition, through the embodiment of the present invention, RREP transmission using route cache information can be basically suppressed. In addition, nodes in the network are monitoring their buffering value Qlen and define a predetermined buffer threshold Qthreshold to compare the difference with the current specific buffering value. There is an interface queue on the MAC layer, which basically operates in FIFO (First In First Out). Therefore, the buffering value may mean the number of packets buffered in the queue.

Hereinafter, it will be described in more detail with reference to FIG.

2 is a flowchart of a method for providing load balancing using path cache information when congestion occurs in a predetermined node according to an embodiment of the present invention.

First, when the source node transmits a RREQ packet to establish a routing path to the intermediate node during the path discovery process, the intermediate node that receives the RREQ packet determines whether the destination of the RREQ packet is itself (S201, S211). . Here, the intermediate node may determine whether the node is the (final) destination node by searching the path information of the destination node in the route cache.

If the destination of the RREQ packet is itself, the intermediate node transmits the RREP packet to the source node (S221).

If the destination of the RREQ packet is not itself, the intermediate node searches the path cache to determine whether there is path information to the destination node (S231).

If there is no route information to the destination node, the RREQ packet is broadcast to the neighbor node (S241).

However, if there is path information to the target node, the intermediate node compares the buffering value (buffer length) and the buffer threshold to determine whether to transmit the RREP packet to the source node (S251). Here, the buffering value may preferably be a queue length buffered through the interface of the mobile node.

Here, if the buffering value does not exceed the buffer threshold, the intermediate node determines that there is no risk of congestion, and transmits an RREP packet to the source node (S261).

However, when the buffering value exceeds the buffer threshold, the intermediate node determines that there is a risk of congestion, and broadcasts the RREQ packet to the neighbor node without performing transmission of the RREP packet (ie, suppression) (S271). ).

Through the above-described process, a path may be established to avoid a node where congestion occurs, and a buffer overflow due to congestion is reduced.

Meanwhile, in the above description, when the buffering value exceeds the buffer threshold, the intermediate node may retransmit the RREQ packet back to the neighbor node instead of transmitting the RREP packet. That is, when the buffering value exceeds the buffer threshold, the corresponding intermediate node may re-broadcast the RREQ packet to the neighbor node to obtain route information to the destination node through the neighbor node and set up routing through another path through the intermediate node. Will be.

In addition, when the buffering value exceeds the buffer threshold value, the RREP packet is suppressed to prevent the use of incorrect path information. In general, the path cache stores the path information that has been successfully transmitted recently. However, the path information also becomes inaccurate as the topology changes dynamically in an ad hoc network with high mobility and large changes in the wireless link. Therefore, the use of incorrect path information causes packet loss and further increases the delay required by rescanning the path. Particularly, when congestion occurs, path discovery failures cause greater overhead and reduce network performance. Therefore, according to the present invention, it is desirable to obtain the latest path information through retransmission of RREQ packets to neighbor nodes. Do.

On the other hand, if the path passing through the node that suppresses the RREP packet is unique and the information of the path cache is correct without error, the performance may be reduced through the above-described method. Particularly, when a packet supporting high quality of service (QoS) and a high priority packet need to be processed, a fast response is requested using path cache information. Accordingly, for this purpose, a QoS flag including priority information may be added to a predetermined field of the RREQ packet. That is, when the RREQ packet includes priority information for providing QoS, the intermediate node determines whether to transmit the RREP packet to the source node according to the priority information. For example, when the QoS flag of the RREQ packet is set to 1, which is the highest priority, the intermediate node determines that the information of the path cache is correct, and transmits the RREP packet to the source node. In this case, it is not necessary to apply the above procedure to the above-described buffer threshold value for the corresponding RREQ packet. Therefore, in step S251, when the QoS flag for providing QoS is set in the RREQ packet, the intermediate node may transmit the RREP packet to the source node without going through a buffer threshold comparison process.

In addition, in the present invention, the RREQ packet may be selectively dropped according to the priority information of the QoS flag. That is, when a buffer overflow occurs due to congestion, a packet having a QoS flag set may not be dropped compared to a general packet so that the node may participate in the path discovery process. And, the packet having a high priority in the priority information of the QoS flag may not be dropped compared to the packet which is not. In addition, when the QoS flag is set as described above, the comparison with the buffer threshold may be omitted, and when the path information for the destination node is included in the path cache, the RREP packet may be transmitted to the source node.

Meanwhile, the priority information of the QoS may be flexibly changed according to a user's policy. A representative QoS technique is Diff-Serv (Differentiated Services), and the Diff-Serv method may be applied to the present invention. For example, in case of using Diff-Serv, if the packet priority is divided into 3 grades of Best Effort, Intermediate, and Emergence, each service may be represented by bit information such as 00, 01, 10, and so on. The bit information may be included in a routing header. If it is dropped due to congestion, a packet that is received later in the FIFO method among the Best Effort packets may be dropped. If an additional drop other than the Best Effort packet is needed, an Intermediate class packet may be dropped in a FIFO manner. Therefore, by applying this order, it is possible to minimize the drop of packets requiring real time.

Hereinafter, the present invention will be described in more detail with reference to FIG. 3.

3 illustrates an example of a method for providing load balancing using path cache information when congestion occurs in a predetermined node according to an embodiment of the present invention.

For example, when node S 310 (source node or transmitting node) attempts to send a packet to node D 320 (destination node or receiving node), node S 310 broadcasts the RREQ packet to a neighbor node. Cast If the node E 311, which is a neighbor of the node S 310, receives the RREQ packet and includes path information destined for the node D 320 in the path cache, the node E 311 compares the buffering value with the buffer threshold value. It is determined whether to send the RREP packet to the node S (310). If the buffering value exceeds the buffer threshold (ie, the threshold of the queue has exceeded), node E 311 rebroadcasts the RREQ packet to the neighbor node. If the node F 312 receives the RREQ packet sent by the node E 311, and the node F 312 includes the path information of the node D 320, the node F 312 responds with an RREP packet. Of course, the buffering value of node F 312 should not exceed the buffer threshold here.

Meanwhile, the RREQ packet of the node S 310 may reach the node G 314 via the node C 313. Here, the node C 313 may broadcast the RREQ packet to the neighbor node because the route information for the target node does not exist. If the node G 314 receives the RREQ packet, and knows the path to the node D 320, the node G 314 transmits the RREP packet to the node S 310.

In the foregoing description, assuming that the node G 314 and the node F 312 are the same distance with the node S 310 at two hops, but the node E 311 is congested, the physical channel states are the same. RREP packets from node F 312 to node E 311 have a greater delay.

Accordingly, the node S 310 sends the packet to the node G 314 using the RREP packet first received from the node G 314 and is routed through the node G 314.

 Although embodiments of the present invention have been described above with reference to the accompanying drawings, those skilled in the art to which the present invention pertains may implement the present invention in other specific forms without changing the technical spirit or essential features thereof. I can understand that. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive.

1 illustrates a path setting method between a conventional source node and a destination node.

2 is a flowchart of a method for providing load balancing using path cache information when congestion occurs in a predetermined node according to an embodiment of the present invention.

3 illustrates an example of a method for providing load balancing using path cache information when congestion occurs in a predetermined node according to an embodiment of the present invention.

<Explanation of symbols on main parts of the drawings>

310: source node

311, 312, 313, 314: intermediate node

320: destination node

Claims (8)

(A) the first node transmitting a Route Request (RREQ) packet for establishing a routing path to the second node; (B) the second node receiving an RREQ packet and determining whether the second node is the destination node; (C) determining, by the second node, whether the route information to the target node exists in the route cache when the second node is not the target node; And If there is route information to the destination node, the second node compares a buffering value, which is the number of packets buffered in the queue, with a predetermined buffer threshold, and sends a Route Reply (RREP) packet to the first node. (D) performing load balancing by determining whether to transmit or not, and performing load balancing. The method of claim 1, The step (d) includes the step of broadcasting, by the second node, an RREQ packet to a neighbor node without transmitting an RREP packet to the first node when the buffering value exceeds the buffer threshold. Load balancing method using path cache information. The method of claim 2, The step (d) further comprises the step of the second node sends a RREP packet to the first node, if the buffering value does not exceed the buffer threshold, load balancing method using path cache information. The method of claim 1, The step (b) includes the step of transmitting a RREP packet to the first node, if the second node is the destination node, load balancing method using path cache information. The method of claim 4, wherein In the step (c), if there is no path information to a destination node in the path cache of the second node, the second node includes broadcasting a RREQ packet to a neighbor node. Balancing method. The method of claim 1, The second node includes a QoS flag that provides information on a priority of a quality of service (QoS) in an RREQ packet, and includes the buffering value and the buffer when the path cache includes path information about a destination node. A method of balancing load using route cache information, wherein the RREP packet is transmitted to the first node without undergoing a comparison with a threshold. The method of claim 6, When the second node includes a QoS flag that provides information on the priority in the RREQ packet, the second node does not drop the high-priority RREQ packet as compared to other packets so as to set a routing path for the corresponding RREQ packet. A load balancing method using path cache information. The method of claim 1, And when the first node receives RREP packets from a plurality of nodes having the same hop number, establishes a routing path using the RREP packet that arrives first.

Images