| | STATEMENT OF INTENT: Enable sufficient end-to-end (E2E) service level performance to critical mission information flows that transit mobile networks comprised of multiple heterogeneous RF links.
| Objective: | Develop technologies for implementing real-time performance monitoring of end-to-end flows which transit multiple RF links.
| Description: | The Department of Defense (DoD) is engaged in efforts to develop an IP-based Airborne Network (AN) which interconnects airborne platforms and provides interconnectivity with space and terrestrial networks. This AN will provide the DoD with the capability to extend mission information threads beyond the original constraints imposed by legacy (non-IP) data links, but its development involves many technical challenges. For additional information regarding the DoD’s efforts to develop this IP-based Airborne Network, please see reference [1]. While this topic description addresses one key technology area associated with the Airborne Network – namely, end-to-end performance monitoring/management – ESC’s HERBB Airborne Networking web site provides the full description of the Airborne Network architecture and its technology challenges. The AN must deliver sufficient end-to-end (E2E) service level performance to the mission threads it supports. Performance monitoring is critical to this task. Similar to terrestrial networks, AN performance measurements may include throughput, latency, error rate, jitter, and availability. However, in contrast to the relatively benign nature of terrestrial wire-line networks, developing a performance monitoring capability for the AN has many challenges: Heterogeneity – AN subnets will employ different RF waveforms. Dynamic, time varying links - link performance will vary due to platform mobility and use of wireless medium. Dynamic network topologies - nodes will enter and exit the network, and aircraft mobility will increase incidence of link degradation/breakage. One of the chief objectives for AN performance monitoring is accuracy. Measurements must accurately reflect E2E performance despite dynamic topologies and time-varying wireless links. Another important requirement is efficiency. Constrained-bandwidth RF links demand that the performance management solution produce minimal overhead. Today’s commercial approaches to performance monitoring typically include device-based management, active probing, and/or flow-based monitoring. Assumptions implicit in the design of these approaches – that networks will be “well behaved”, relatively static, with ample bandwidth -- result in inadequate solutions for AN performance monitoring for the following reasons. Device-based management - the polling approach to data collection tends to be inefficient and adds costly overhead. Additionally, the centralized nature may place inordinate processing demands upon the monitoring server due to frequent network state changes. Finally, fusing atomic-level measurements of each device into a view of end-to-end performance is likely infeasible in dynamically changing networks. Active probing - the sampling mechanism may not be sufficiently responsive in dynamically changing networks: the more frequent the network changes, the more frequent the probing to maintain an accurate, up-to-date view of changing performance, resulting in costly overhead. Flow-based monitoring - flow-based monitoring strategically places passive monitors at aggregation points within the network; however, in the airborne network, aggregation points will change with node mobility, dynamic node membership, and link topology changes. The viability and efficacy of flow-based monitoring under a dynamic, meshed network topology has not been evaluated. Innovation is required to develop a new approach to performance management of RF networks such as the Airborne Network. Innovative approaches may adapt some positive aspects of active probing, device-based management, or flow-based management, or alternatively, develop an entirely new approach to performance monitoring.
| | PHASE I: Identify and/or develop potential approaches for conducting E2E network performance monitoring in a dynamic mobile environment comprised of heterogeneous RF networks. Present the pros and cons of candidate solutions. Use simulation or prototype techniques and analyze performance and robustness of solution(s) to determine feasibility.
| | | PHASE II: Complete the design, develop, and demonstrate a prototype that conducts E2E performance monitoring of information flows across a dynamic mobile set of heterogeneous RF networks.
| | DUAL USE COMMERCIALIZATION: Military application: E2E performance management of information flows across critical C2 networks is a requirement present in a large number of military networks, both airborne and terrestrial based. Commercial application: E2E performance management is a broad technical issue that cuts across almost all networks, military and commercial, wireless and wired, to one extent or another [2], [3].
| References: | 1. ESC HERBB Airborne Networking web site, http://www.herbb.hanscom.af.mil/Hot_Buttons/Airborne_Networking/index.htm.
2. MIT’s Technology Review Magazine, http://www.technologyreview.com/read_article.aspx?id=14407&ch=infotech
3. Airborne Internet Consortium, http://www.airborneinternet.org/
| | Keywords: | Airborne Network, Airborne Internet, Mobile and Ad-hoc Networking, Network Modeling and Simulation, Performance Management |