The spider and the cellphone

A sidebar to the Bell Labs special report, Reviving an icon

Unlike his colleagues in Murray Hill, Bell Labs technical staff member Sharad Ramanathan does his research in Boston, where he has access to Harvard University’s biology labs. There he watches spiders weave their webs and the embryonic cells of mice divide and grow. At first, this research would seem to bear no relation to communications technology, but Ramanathan is exploring how the biological network can be applied to the communications network.

A spider can build its web without the benefit of perspective. It can’t look down from above at the geometric design it is constructing. Instead, Ramanathan believes the spider can measure the tension at any given point on the web to determine its overall shape.

The same goes for the cells dispersed through the various fetal limbs of a mouse. They manage to grow legs in exact size and proportion to the mouse’s other limbs though they have no direct communication with the development cells in the other limbs. Why can’t these principles be applied to a communications network, Ramanathan asks.

Instead of a monolithic central intelligence controlling the complexity of thousands nodes on the network, through the algorithms and architectures that Ramanathan and his collaborators are working on each node could gather information from just its neighboring nodes and infer the status of the entire network. The result could be an ‘aware,’ self optimizing network that could support millions of nodes, all making individual routing decisions based on local information.

Such a network could be scaled to incredible size and dispel the notion of a distinct core and distinct endpoints in a communications system.