Integrating Directional Links with Omni-Directional Tactical MANET Stack: An Experimental Analysis

Utilizing directional links can augment the performance of omni-directional tactical MANETs. In this paper, we report the results of a modeling and simulation study focused on capturing the performance enhancement of tactical MANETs measured in terms of throughput and delay. We compare the performance enhancements of two alternatives for the integration of directional capability with omni-directional tactical MANET stack. While the first alternative utilizes a standard layered MANET stack, the second alternative takes advantage of a modified stack that can cope with the undesired effects of mobility by inserting an extra layer referred to as mobile layer between the data link layer and the network layer. Consequently, the standard and modified stacks integrate the directional capability at the network and mobile layer, respectively. Our findings show that integration at the mobile layer provides significant throughput and delay improvements and as such is the preferred way of adding the directional capability to omnidirectional MANET stack. Based on our results and in conjunction with the mobility mitigation characteristics, we conclude that the modified stack represents a better alternative than the standard stack for use with military programs such as JTRS GMR and FCS

I. Introduction

As evidenced by the works of [13], [14], [16], [18], and [19], augmenting the performance of omni-directional networks with a directional capability has been the subject of great interest. In the case of tactical ad-hoc networks, the use of directional capability can greatly decrease the interference effects, increase the achievable signal-to-interference ratio, increase the achievable throughput, lower end-to-end latency, and strengthen jamming tolerance. However, the high degree of mobility and the lack of a centralized controlling entity introduce serious challenges in fully utilizing directional capabilities in such omni-directional networks.