Article Preview
Top1. Introduction
MANET (Mobile ad hoc Network) is a self-organized, infrastructure less wireless network without any centralized administration stated by (Murthy & Manoj, 2004). The goal of QoS aware routing is to identify the optimal path that satisfies the stringent requirements of bandwidth and delay for real time applications. Multimedia applications such as audio and video have much more stringent QoS requirements. For a network to guarantee QoS deliveries, it has to reserve and control resources like minimal bandwidth requirement and delay guarantees. The main issues and challenges of MANET routing protocols is to deal with link failures and route recovery in these situations. The efficiency of route recovery affects the overall performance of MANETs. AODV has two route repair strategies to deal with link failure. Routes are repaired by either reestablishing a new route starting from source node, or it can be locally repaired by the node that detects the link break along the end-to-end path proposed by various authors of Zhu et al. (2006), Crisostomo et al. (2004), Qu et al. (2009), Zhao et al. (2010) and Veres et al. (2001) in their research.
Source initiated route recovery will lead to better performance but in other situations local repair will be the more appropriate choice. In either case, the route update process attempts to use flooding-search to obtain alternate path to the destination, it results in the decrease of the network throughput as well as long delays. The problem gets worst when mobility is high. Our approach consists of prediction of link failure (Li, Liu, & Jiang, 2008) of next node link by signal strength detection technique with few additional control messages, the predecessor node searches an alternate route utilizing a mesh structure and establishes a path after receiving an acknowledgement from an alternate node.
1.1. Challenges in QoS Routing
QoS routing has to consider application requirements and the availability of network resources. Consequently, QoS routing in ad hoc networks exhibits great challenges which are reflected in Table 1.
Table 1. | Dynamic topology | The topology of wireless ad hoc networks changes with time due to mobility scenario. It leads to imprecise network state information at the nodes therefore it is a challenging task to provide QoS guarantees. |
| Shared bandwidth | The bandwidth in a wireless network is shared by other nodes in the network. The transmission from a node not only consumes local resources but also consumes the bandwidth of the neighbors within the contention range. |
| Resource constraint | Mobile nodes have power limitations as they are battery operated. If the battery power is not efficiently used, mobile nodes will fail quickly which will affect the network availability and functionality. |
| Lack of centralized control | The absence of any centralized control demands the routing protocols to be self-creating and self-organizing. Further, the protocols must also be distributed in nature. |
Multimedia application is one of the most promising services over the Internet. Portable devices in MANET such as PDA and Laptops have different access patterns. The bandwidth requirement for the multimedia applications with desired level of Quality of service is one of the major challenges for the future wireless networks generation. Multimedia applications require a greater amount of bandwidth (Varaprasad & Wahidabanu, 2011) therefore a large number of data packet flows require many nodes to participate in the transmission. Hence, communication overhead and delay latency increases. The proposed protocol considers delay and bandwidth at every node while considering route discovery. The rest of the paper is organized as follows: Section 2 describes related work about QoS routing and link failure recovery schemes. Section 3 presents our proposed QoS routing protocol. Section 4 remarks the analytical model of our proposed approach. Section 5 reflects our simulation results. Finally, in Section 6 we conclude the paper with directions of future work.