A performance cost/benefit analysis of adaptive computing in the tactical edge

Abstract

Tactical Edge Computing is a promising solution to address the challenges of processing and managing large volumes of data collected by sensors deployed at the tactical edge. Tactical edge networks often lack sufficient bandwidth to transfer data at high rates. Much of the sensor raw data might be uninteresting and would waste networking resources to transmit. Edge computing solves these problems by staging the processing capabilities for sensor data at the edge, close to the data source. Local data processing reduces bandwidth needs in disadvantaged tactical networks. Furthermore, it could reduce the latency of data processing and avoid congestion in the tactical network. However, static solutions that deploy edge services could also be problematic because of the unpredictability of the tactical Edge – the source / location of the data may not be known a priori, and there could be significant changes, such as sensors and nodes going offline. Therefore, tactical edge computing solution must be adaptive, where services are deployed on demand based on the sensors tasked and the mission requirements. This paper presents the work of the NATO IST-193 work on adaptive tactical edge computing and its analysis on the adaptivity benefits and overhead costs involved.