An Adaptive Geographic Internet Sub-Protocol for Dynamic Wireless Networks Navy SBIR FY2005.2 Sol No.: Navy SBIR FY2005.2 Topic No.: N05-108 Topic Title: An Adaptive Geographic Internet Sub-Protocol for Dynamic Wireless Networks Proposal No.: N052-108-0105 Firm: Intelligent Automation, Inc. 15400 Calhoun Drive Suite 400 Rockville, Maryland 20855 Contact: Hongjun Li Phone: (301) 294-5275 Web Site: www.i-a-i.com Abstract: In order to seamlessly integrate ad hoc and sensor networks with the Internet, we introduce a novel entity called Internet Sub-protocol Manager (ISM) that serves as the "bridge" between the standard IP and our geographic sub-protocol (SP). The standard IP stack (either IPv4 or IPv6) needs not to be changed. Rather, the ISM will encapsulate the original IP packets within the sub-protocol payload and provide an IP-in-SP tunneling. The sub-protocol exploits the geographical and mobility information available from the GPS equipped communication node and achieves position-based routing that incurs less storage and less processing overhead. The geographic routing provides alternate paths for such wireless nodes. While current geographic routing protocols are impractical, our proposed Cross Link Detection Protocol (CLDP) guarantees that geographic routing between all pairs of nodes will never fail. We have demonstrated that our approach achieves geographic routing correctly on real radio networks. ISM and our geographic routing protocol can be easily configured and we provide an efficient mechanism for wireless routing packet format discovery. In addition, we propose to adaptively manage the routes via proactive prediction of link expiration time and describe a method for differentiating link failure and congestion. Benefits: The proposed architecture, Internet Sub-protocol Manager, and the practical geographic routing protocol provide an innovative and seamless way for integration of ad hoc and sensor networks with the Internet. Our scheme significantly increases the practicality of position-based routing and achieves integrated, efficient and robust networking for the military communication needs, including war-time command and control, surveillance and control, and homeland security, etc. The market for military applications is quite large. Other potential commercial applications include border and coast patrol, law enforcement agency, emergency control center, and various civil applications, possibly with huge amount of users. The size of this market is quite large and hard to estimate. In particular, the increasingly popular wireless sensor networks will certainly benefit from our work and they provide great opportunities for commercialization. We expect the aggregate market size will be similar to or larger than that of military applications. Return