The K-Neigh Protocol for Symmetric Topology Control in Ad Hoc Networks

Abstract

We propose an approach to topology control based on the principle of maintaining the number of neighbors of every node equal to or slightly below a specific value k. The approach enforces symmetry on the resulting communication graph, thereby easing the operation of higher layer protocols. To evaluate the performance of our approach, we estimate the value of k that guarantees connectivity of the communication graph with high probability. We then define k-Neigh, a fully distributed, asynchronous, and localized protocol that follows the above approach and uses distance estimation. We prove that k-Neigh terminates at every node after a total of 2n messages have been exchanged (with n nodes in the network) and within strictly bounded time. Finally, we present simulations results which show that our approach is about 20% more energy-efficient than a widelystudied existing protocol.

A Topology Control Algorithm for Interference and Energy Efficiency in Wireless Sensor Networks

“…( r : (r ∈ RN brsp) ∧ (power(p, r) > power(p, q)))} highestp ← any q such that q ∈ highestSetp 4 myP owerp ← power(p, highestp) 5 The set P os p is the set of 2-tuples q, x q , y q , where x q and y q are the Euclidian x, y-coordinates of process q. We do not specify a particular way for p to obtain G 2h p and P os p as the focus of this paper is on the topology generated by the algorithm (they can be easily obtained by the exchange of messages with topology information).…”

“…We compared TCO with GG because GG presented good results in [6] for an interference metric similar to ours. In particular, it outperformed the algorithms CBTC [21] and KNeigh [5]. Finally, we compared TCO with RNG because it is another classical result of computational geometry upon which many topology control results are based.…”

Self Organization and Self Maintenance of Mobile Ad Hoc Networks through Dynamic Topology Control

“…The topology control problem [22] has been deeply investigated in the literature in recent years, including theoretical studies aimed at characterizing optimal topologies according to some performance metric (see, e.g., [19,21,25]), and more practical approaches presenting distributed, localized topology control protocols, sometimes with proven performance bounds with respect to optimal [2,3,8,10,15,18,20,27,28]. Included in this work is our original k-Neighbors approach [2,3], upon which this current work builds.…”

“…Included in this work is our original k-Neighbors approach [2,3], upon which this current work builds. Some papers [4,5] also addressed the topology control problem with the goal of reducing interference, instead of energy consumption as traditionally done in the topology control literature.…”

“…The ultimate goal of our topology control protocols is to cause N s (u) to contain k (or slightly more than k) nodes, where k is an appropriately chosen parameter 5 . Motivations for our interest in the number of symmetric neighbors of a node can be found in [2].…”