Disruption/Delay tolerant networking (DTN) is a promising solution to mitigate the effects of periodic link outages that last seconds to minutes, which are prevalent in high capacity airborne tactical networks that are directional in nature and highly affected by aircraft body blockage. There are several considerations for applying the DTN Bundle Protocol (RFC 5050) in tactical edge networks which have ciphertext/plaintext boundaries resulting in multiple deployment options: 1) black-side DTN tunnel, 2) red-side DTN interface, 3) red-side DTN proxy, and DTN-enabled applications. Each of these approaches have benefits and limitations. In this paper, we present a DTN proxy implementation that can work on both red or black sides of tactical edge networks and show test results that help us better understand the capability of the approach, its limitations, and in what circumstances it is most appropriate. Through experimental and theoretical analysis, we show that TCP file transfer speed can be increased by up to 143.39% and 218% for two-hop and three-hop aerial backbone networks respectively when each link suffers 25% independent outage.

In recent years there has been a strong desire in the U.S. Department of Defense to augment traditional ground and satellite communications with a high capacity aerial tier. High capacity airborne links are often directional in nature and highly affected by aircraft body blockage, exhibiting unique outage characteristics compared to ground or satellite networks. To mitigate the effects of periodic link outages that last seconds to minutes, disruption tolerant networking (DTN) technology has been proposed. In this article we examine applying the DTN Bundle Protocol (RFC 5050) to ship-to-shore networks for traffic flowing over the aerial nodes. Specifically, we examine applying DTN proxies, tunnels, and interfaces to both the plaintext and ciphertext side of military networks to understand architecture and design considerations and limitations.

In a tactical environment, disparate surface nodes are often unable to communicate effectively with each other through line-of-sight radios due to long distances and terrain blockages. The Joint Aerial Layer Network (JALN) consists of a set of unmanned aerial vehicles (UAVs) that provide surface range extension to surface nodes while also providing reachback to the Global Information Grid (GIG). These capabilities are achieved through a downward facing antenna that provides surface-to-surface communications relay, as well as directional, high data rate crosslink antennas that connect the UAVs to each other. In this paper, we examine the implications of platform dynamics on crosslink availability and up/downlink coverage area in JALN. We study the availability for various orbital patterns and show that smaller orbits provide a larger high-availability coverage area. However, crosslink and up/downlink blockage is more likely during aircraft banking, which should be taken into consideration when designing an orbit. In addition to the orbital patterns, we also study how crosslink availability is affected by antenna placement. A single wing-mounted pod design suffers greatly due to field-of-regard restrictions. We compare the single active pod design with other options to improve performance including two active pods and body-mounted antennas.

The quality of the scheduling has a strong impact on the overall application performance because of process and data affinities. However, this issue is now becoming critical due to the variable memory access latencies in NUMA (Non-Uniform Memory Access) architecture, because in NUMA architecture local data access being significantly faster than remote access, then data locality emerges as a critical criterion for scheduling threads and processes, and it becomes important to be able migrate memory together with their accessing tasks.To perform memory migration, we present memory migration on-demand policy to enable automatic dynamic migration of pages with low cost when they are actually accessed by a task. We use PTE flag setup with the help of madvise system call and the corresponding Copy-on-Touch code added in the page-fault handler which allocates the specific page near the accessing task.

With the increasing size and complexity of emerging military communication networks, low footprint emulation methods are required to evaluate protocols, algorithms and user performance. In this paper we describe a new emulation architecture that is scalable, modular, and responds to real-time changes in topology and link characteristics. We use this scalable emulation testbed to evaluate system performance in a simple Satcom network topology, and enhance it to emulate Dynamic Resource Allocation using distributed software architecture.