The emphasis was on the future at the 2006 MIT Research and Development Conference and LIDS Director Vincent Chan’s presentation on Future Integrated Broadband Fiber, Wireless, and Satellite Networks was no exception.
Professor Chan began his talk by pointing out the different challenges presented by fiber, wireless, and satellite technologies, and closed by offering a vision of the future: an integrated global network, which would utilize interconnected fiber, wireless, and satellite systems, to result in both an increase in data rates and a decrease in cost.
Professor Chan highlighted the current research at MIT before asking a few intriguing questions: “Where is all this research leading to? Will it affect the future, or stay bottled up?” He emphasized that cooperation between researchers and industry professions is essential for the groundbreaking research conducted at MIT to make an impact in the real world, and praised the Research and Development Conference for its ability to “disseminate information.” According to conference sponsors at the MIT Industrial Liaison Program, increased communication between researchers and industry leaders is the precise goal of the conference.
The need to address communications technologies from both cost and performance perspectives was an important emphasis of Professor Chan’s presentation. Because the average American will not pay more than about $50 per month for Internet access, no matter its speed, Professor Chan explained, “a decrease in cost per bit is essential.” In addition to cost, the other challenges inherent to the evolution of communication include the increasing mobility of users and the very nature of modern Internet traffic. Professor Chan described the problems facing the mobile user, who may be connecting “over the ocean, with no previously deployed infrastructure.” Inclement weather can impact the reliability of transmissions; for cellular and satellite users, “when the channel fades, the data transaction cancels.”
These issues can be frustrating for users. Professor Chan pointed out that “while most computer users are dormant 99.9% of the time, when they send something they want it done quickly.” Also the transactions can have what is known as heavy tail behavior, in which 1% of all transactions contain 99% of the data. These transactions may contain large files such as medical images, or as Professor Chan joked, may be caused by “teenagers downloading movies.” Most data does not travel in predictable predetermined streams; instead, 90% of data traffic is “bursty,” or unscheduled. Hence, researchers and industry professionals are faced with the problem of “random multiple access,” the difficulty of predicting when users will need bandwidth, and how much bandwidth they will need.
While Professor Chan acknowledged that there are “realities of the channel that you can’t duck,” he suggested that problems of stability, scheduling, and low data rates can be addressed by an integrated network structure. He argued that solving these problems will require a new and different architecture, but stressed that “this architecture needs to remain current for multiple generations of technologies” to accommodate changing technologies in a cost-effective way. In closing, he charged his listeners with the task of solving the problems he had presented, and emphasized that the challenges of the future must be met with innovative and economic solutions.