Research: Highlights

Bookmark and Share Email Print

04.09.09 - New Grant Supports MANET Research

Yousefi'zadeh is setting up a test bed for the 
software-defined radios in the Calit2 
Networking lab.

 


When it was first introduced, wireless communication required a supporting infrastructure – a cell tower, a router connected to a wired network or an access point on a local area network.

The introduction of MANETs (Mobile Ad hoc NETworks) eliminated that constraint. MANETs are collections of autonomous mobile wireless nodes that can transmit and receive a variety of data and rich media content in real time without relying on established infrastructure.

Now, a new three-year, $1.5 million grant from the Boeing Co., administered by Calit2, is funding research to further develop a MANET testbed. Principal investigator Homayoun Yousefi’zadeh and co-principal investigator Hamid Jafarkhani are creating custom MANET software that controls software-defined radios (SDRs). The electrical engineering and computer science professors are designing and implementing new protocols in an attempt to improve MANET performance and efficiency, while reducing costs.

MANETs are most useful when flexibility and fluidity matter, says Yousefi’zadeh.  Because they are independent of fixed infrastructure, MANETs can be constructed for a specific task, disassembled when the job is complete, and reassembled elsewhere, making them practical for military, emergency, and even sensor applications.
Currently though, they have limitations.

Overcoming Restrictions
Physical obstacles between two nodes can impede transmission, so existing MANETs typically rely on line-of-sight links to ensure quality reception. In addition, interference caused by simultaneous transmissions can limit the available bandwidth, impacting the types of content delivered over such networks.

“If there are multiple users on the network at the same time, the amount of bandwidth that can be dedicated to an individual session will be low,” says Yousefi’zadeh. “How do you schedule everybody [so their communication is received] in real time and still give them decent quality?”

He believes the answer is cross-layer protocol software that allows for the exchange of information among adjacent layers comprising the network protocol stack. Data is generated at the application layer and moves from layer to layer as it is processed, formed into packets and transmitted. 

Glitches along the way can impact delivery and quality.  For example, a mobility issue that temporarily affects the quality of the wireless links carrying messages might be eliminated by a wait of milliseconds. Real-time communication requires making continuous routing decisions that don’t always take into account link-quality issues. But “cross-layer exchange of information” gives the network layer access to link quality information so it can contribute to the final routing decision. 

Adapting to Dynamics
The “smart” radios, researchers hope, can automatically adjust to changing circumstances. Communication in a jungle, with a lot of bushes and not many lines of sight, would necessitate a different mode of operation than a flat surface, Yousefi’zadeh says. “Programming this type of radio with software allows you to pick your mode of operation dynamically, as the scenario of operation changes.”

The software will also facilitate scalability by letting devices change the way in which they transmit or receive messages based on the number of nodes participating in a communication scenario.

The hardware and software act in tandem, giving the user options tailored to optimize performance in specific environments.  “You have to [make available] a limited number of modes that are automated to perform best in certain kinds of environments,” Yousefi’zadeh says. “If I’m on a flat, open field, I choose mode 1; if I’m in a scattering environment, I use mode 2.”

Increased sophistication brings more complex architecture dilemmas. In infrastructure-supported wireless communication, cell towers and routers play a dedicated role: they relay messages between nodes. But in MANETs, all nodes in the network serve as both routers and end nodes, making routing protocols more complicated.

Each node – a radio or other device – gets instructions about whether it’s supposed to relay or decrypt the packets it obtains. “Each node can relay information to other nodes but it might have its own content, too,” Yousefi’zadeh says. “Part of our work is to improve the performance of current systems. We’re hoping our solution will be better- performing, more scalable and less expensive, yet still secure.”

It all comes down to getting the highest amount of bandwidth and connectivity in mobile communication networks while using minimum battery power. The networks must maximize security and scalability, while considering dynamic environmental and timing aspects.

Each MANET function has its own requirements, further complicating the challenge. For example, audio and video might be subject to time constraints but require less in the way of detail, while a surveillance photo requires extensive detail, but perhaps less attention to delay issues. “If you have a collection of these applications and everything is dynamically varied, secure operation becomes very complex and challenging,” Yousefi’zadeh says.

-- Anna Lynn Spitzer