Logical clusters in a DHT-Paradigm for scalable routing in MANETs

Abstract

Connectivity of nodes in the logical network is the minimal requirement for the functionality of a distributed hash table (DHT)-based routing protocol. In case of high mobility, frequent changes to the network topology introduce higher maintenance overhead for updating the mapping information that worsens not only the performance of these protocols, but also restricts their pertinence only to the networks with low mobility. In addition, the route to the destination node in such protocols is not immediately available which result in high lookup latency for the requesting node.

In this paper, we proposes a novel 3-dimensional logical cluster-based DHT routing protocol for mobile ad hoc network (MANETs) that use logical clustering and an effective replication strategy to reduce the routing overhead and lookup latency introduced by the above problems. Simulation results substantiate the effectiveness of the proposed protocol in terms of reducing the delay by slightly increasing the control overhead.

Introduction

The Distributed Hash Table (DHT) structure has emerged as a scalable substrate to provide a diverse set of functionalities, like location-independent identity, information distribution, and location service [1], [2] in various self-organized applications or systems over the Internet. It has come forth as a pragmatic additional technique to the design of spontaneous and self-organizing networks.

In the past few years, DHT has been adopted for routing in large scale mobile ad hoc networks (MANETs) by implementing it directly at the network layer, e.g. in [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. The core objective of such approaches is to handle the scalability issues in MANETs by eliminating network-wide flooding. DHT-based approaches outperform non DHT-based approaches when the number of nodes, or the query rate increases, since the later do not avoid flooding in the network [4], [5], [8], [10], [18], [19]. Various application scenarios are listed in Table 1, where DHT-based routing protocols would be helpful in improving the communication among network users.

Handling the mobility of nodes has always been a major challenge in ad-hoc networks. DHT-based routing protocols for MANETs do not support high mobility. Because a logical network is maintained over the physical topology in DHT-based routing in which each node computes a logical LID in addition to its universal identifier (UID), i.e., IP/MAC address. In case the neighborhood of a node P changes or P moves to another place, P re-computes its LID that reflects its relative position with respect to its new neighborhood. In case of high node mobility, the frequency of node's LID re-computation increases which would increase traffic overhead, computation overhead, packet collision and chances of network inconsistency. This would generate more traffic to update the network status and subsequently increase end-to-end delay in the network. These problems need an immediate attention and require an optimal solution. Table 2 illustrates the definitions of few important terms related to DHT-based routing in MANETs.

In this paper, we propose a novel 3-dimensional logical cluster-based DHT routing protocol for MANETs (3D_cRP) that introduces logical clusters in the network in order to develop an efficient mechanism that handles the mobility of nodes. The advantage of using logical clustering in 3D_cRP is twofold. i) It would directly reduce the number of update messages produced at each node in the network; ii) The idea of logical clusters would also minimize the impact of the mismatch problem [21] when sending a packet from a source node to a destination node, resulting in a low path stretch penalty that minimizes the end-to-end delay and routing overhead. In addition, the proposed protocol introduces a replica management strategy that effectively replicates the mapping information at various nodes, which would help to minimize the lookup latency and lookup traffic.

• To the best of our knowledge, 3D_cRP is the first logical cluster-based DHT routing protocol that is unfolded directly at the network layer for routing purpose. None of the existing DHT-based protocols [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16] have used logical clusters: i) to handle the mobility of nodes; ii) to deploy a replication strategy; iii) to reduce the path-stretch penalty.

• In 3D_cRP, only logical cluster heads (LCHs) computes their LIDs that reduce the impact of node mobility. As a result, all the member nodes in the logical cluster would be able to move freely and those member nodes do not require LID re-computation in case the neighborhood changes or any other member node joins/leaves the logical network.

• In 3D_cRP, we use Shepard's interpolation method [22] with some modification to compute the LID of a LCH relative to its neighboring LCHs.

We assume a self-organizing communications scenario, where mobile handsets are used as ad-hoc network nodes, both for data storage and intelligent relaying in the absence of base stations. To control stability, re-configurability and scalability, self-organized ad-hoc networks communicate via multi-hopping over relay nodes, which are clustered into various groups.

3D_cRP adopts the cluster formation algorithm similar to the one proposed in CBRP [23], but unlike CBRP, 3D_cRP mainly focuses on using logical clusters and LCHs: i) to avoid flooding in the lookup and routing process; ii) to reduce the impact of mobility in order to reduce the control overhead; iii) to replicate the mapping information at different cluster heads in order to reduce the lookup latency; iv) to reduce the loss of packets in case an anchor node moves/fails. To the best of our knowledge, none of the existing DHT-based routing protocols has used logical clustering to attain these objectives.

The rest of this paper is organized as follows: In Section 2, we provide a review of the existing protocols. Section 3 outlines the problem statement. The motivation behind the proposed 3D-structure and the detail of the proposed DHT-based routing protocol in MANETs is explained in Section 4. Simulation results are presented in Section 5. Section 6 concludes the paper.