A Routing Protocol for Delay-Sensitive Applications in Mobile Ad Hoc Networks

Abstract

Generally speaking these protocols are classified into two broad categories: proactive (table driven) and on demand The Ad hoc On Demand Distance Vector (AODV) routing routing (source initiated) [4,5,6]. Proactive protocols are less protocol is intended for use by mobile nodes in ad hoc efficient than the later, as they need more network resources networks. To provide quality of service, extensions can be to maintain an up-to-date routing information to every other added to the messages used during route discovery. These node in the network. On the other hand, on-demand routing extensions specify the service requirements which must be met protocols initiate route discovery (route request and route by nodes rebroadcasting a route request (RREQ) or returning maintenance) as and when required and these protocols only a route reply (RREP) for a destination. In order to provide maintain the part of routing table currently in use. On demand quality delivery to delay sensitive applications such as voice routing protocols react to topology change quickly and also and video, it is extremely important that mobile ad hoc save routing overheads compared to proactive routing networks provide quality ofservice (QoS) support in terms of protocols. Many routing protocols have been proposed for bandwidth and delay. In spite of using IEEE 802.11 as MANETs. Most of the current ad hoc routing protocols only medium access control (MAC), most of the ad hoc routing use number of hops as a measure of route cost in making protocols do not consider contention time which occurs in the routing decisions. But in this some important link capacity medium reservation. Large contention times can be more properties are ignored due to simplicity and ease of critical than hop counts in determining the end-to-end delay. implementation. Firstly, as each node has a different traffic Most existing routing protocols for mobile ad hoc networks load, therefore the average number of packets in the queue are designed to search for the shortest path with minimum and the associated queuing delay at each node is different. hop counts. However, the shortest routes do not always Secondly, the number of neighbor nodes as well as their provide the best performance, especially when there are traffic patterns are different, and thus, nodes that have more congested nodes along these routes. In this paper we propose number of active neighbors may encounter more collisions. If an on demand delay based quality of service (QoS) routing nodes with more traffic fall on the shortest route, it may protocol (AODV-D) to ensure that delay does not exceed a actually introduce longer end-to-end delay, even though the maximum value for mobile ad hoc networks. This protocol number of hops along the chosen route is minimum. will take into consideration MAC layer channel contention Furthermore, if some of the heavy nodes are congested, it information and the number ofpackets in the interface queue may lead to massive packet drop rates and subsequent in addition to minimum hops. MAC l...

QoS Aware Multipath Threshold Routing for Mobile Ad hoc Networks

“…However studies have shown that single path routing based on minimum hop count metrics may not always provide best performance, especially in MANETs when there are nodes along the route which are congested or have less resource such as bandwidth or residual energy [8,9]. In single path routing, the discovered routes may not be reliable because they could easily be broken due to node mobility or unpredictable link conditions, resulting in transmission failures and delay in the network, making the protocol unreliable.…”

Handling Malicious Flooding Attacks through Enhancement of Packet Processing Technique in Mobile Ad Hoc Networks

“…However, the neighbor suppression mechanism does not check whether it receives the corresponding RREP packet or not. Hence, prevention from unreachable destination [11] is not possible. It is also vulnerable to the RREQ flooding attack conducted by the generation of many different source addresses.…”

“…The local density, ζ, of a node is defined as the number of neighboring nodes lying under its transmission range. With respect to local density, Rbuffer(control) and Rbuffer(data) can be defined as in (10) and (11).…”

“…If there is no corresponding RREP packet, the RREQ packet can be used to conduct RREQ flooding attack. On the other hand, the FAP scheme does not check the unreachable destination [11] and determines the threshold as the maximum originating data packets in a time period. However, data packets flooded for conducting the data flooding attack are flooded continuously at a high rate.…”

“…Even though the FAP scheme can protect the RREQ flooding attack by a low threshold, it has path initiation overhead since it cuts off the path with the attack-suspected neighboring node in case of a data flooding attack. When the amount of attack traffic is high, the proposed scheme has higher gain of legitimate packets than the FAP scheme because it prevents from the unreachable destination [11] by serving some of the received data packets in the case of a data flooding attack, rather than totally cutting the path off. Figure 9 shows that the end-to-end packet delivery ratio between the source node and the destination node is reduced as the amount of attack traffic is increased.…”