Capacity planning strategies for net-centric applications

Military communications have several unique characteristics, limitations, and requirements such as need to rapidly deploy wireless ad-hoc networks, links with limited data rates and higher bit error rates, lower link availability, information assurance and covertness requirements, jamresistance, resilient reach-back, network availability, and quality of service (QoS). Traditionally, military networks have employed application-specific ("stovepiped") communications links but there is a significant effort to adopt network-centric architectures based on commercially available and technology. Transitioning to a network-centric architecture requires establishment and management of multiple network layers. Each layer adds its own overhead, and may require data processing and control signaling, adding to the required network capacity. Additionally, information assurance, quality of service, MANET, XML, and Web service protocols and technologies add to the total traffic overhead. Since networks will be shared by multiple users at a given time, the traffic on various segments of the network is expected to be unpredictable and bursty. For achieving an acceptable performance level for unpredictable traffic scenario, the network capacity utilization should not exceed a threshold level. Additionally, network inefficiencies may originate from integration complexities of heterogeneous networks. This paper examines the capacity requirements for a variety of military applications supported by a network-centric architecture. In particular, the overhead and inefficiencies of physical layer, link layer, network layer, transport layer and application layer protocols will be examined. For each of these layers, a variety of commercial and military protocols, that enable net-centric applications, are discussed with emphasis on its impact on capacity planning.