Like most PhDs., my research has meandered through various topics. Here is a brief description of some of the concepts and ideas that I have work on.
Vehicular Ad Hoc Networks (also known as VANETs) are an emerging class of applications of Mobile Ad Hoc Networks. The goal is to enable inter-vehicle information exchange to enable safety messaging, traffic information dissemination and general purpose Internet access. There are several challenges to deploying inter-vehicle communication capability. One of the significant challenges is the unreliability of the wireless communication medium and its impact on safety messaging. The IEEE 802.11p standards body is working to develop WAVE (Wireless Access for Vehicular Environments) to enable vehicle-to-vehicle and vehicle-to-infrastructure communication. Our work focuses on challenges in the system such as the rapidly changing topology of the system as vehicles traverse at high speeds on highways. We haveprepared an analytic model of message propagation that accounts for rapidly changing topology, transmission range of the physical radio and the speeds at which vehicles travel. Furthermore, based on the achievable performance, we are preparing results for the optimal placement of access points in the system for optimal performance while minimizing infrastructure.
Delay Tolerant Networking
Delay tolerant networks (DTNs), also known as opportunistic or intermittently connected mobile networks (ICMNs) are networks where nodes are mostly partitioned and come into intermittent contact with other nodes in the network, by virtue of mobility or sleep-wake mechanism. The existence of end-to-end paths from source to destination nodes is extremely unlikely. Data in the network is forwarded whenever nodes come in contact with each other. Thus, the path-based routing mechanisms are rendered invalid for most of the time. Our work focuses on the achievable performance of message propagation for given phyical conditions such as topology, sleep-wake scheduling mechanisms and mobility in the system.
A semantic network is often used as a form of knowledge representation. It is a directed graph consisting of vertices, that represent objects, and edges that represent semantic relations between the objects. Our research on semantic networking in the context of mobile ad hoc networks focuses on developing a framework for enabling routing based on knowledge in the network. Routing in networks is often independent of the underlying structure. However, there is opportunity to exploit implicit knowledge in the network. This concept is relevant to mobile ad hoc networks and sensor networks that are resource constrained and self-contained. Examples of implicit knowledge include topology or gradients that exist in the network. Sensor Networks are often tasked to collaborate and co-ordinate with other nodes that bear some relationship. Our work proposes the use of attributes or labels that enable the nodes to use this relationship in routing. The attributes can be used to create routing rules or overlays that are used to route data more efficiently.
An example of this concept is location based routing, the location of source and destination is embedded in the packet and at each hop in the network a routing decision is made based on the location attributes. This is analogous to how our current postal system works where the address of the addressee and the location of the post office determines the route of the letter with local decision making built in to the system.