CSD 2018
The Network Tapestry
By Jeff Lawrence
Twenty years ago, I was in a UCLA engineering lab programming an 8080 microprocessor in Assembly language to play a song entitled "Greensleeves." Back then, the network was telephony;
today, engineering students work with 64-bit microprocessors and Java, and use e-mail, Web browsers, and other tools to access worldwide resources in ways that were almost unimaginable 20 years ago. What will the network be like in another 20 years?
First, bandwidth will continue to increase. The transmission backbone will effectively become a pipe based on wavelength-division multiplexing (WDM) and its successor technologies. Electronic switching in the center of the network will move towards
obsolescence, replaced by optical switching. Access to the network will happen via fiber, cable, wire, and wireless. Wireless access will increasingly use spread-spectrum techniques, as radio spectrum allocation standards and regulations catch up to the technology. Intelligent appliances and devices that monitor, control, and interact with the environment will become ubiquitous. These devices will assist people in matters such as personal communications, electronic commerce, personal security, and information
management.
Nanoprocessors - new siblings of microprocessors - will integrate processors, DSPs, RAM, and ROM into a single chip. They will be used extensively for appliances and devices with low power budgets.
Deployment of quantum dot technologies will begin for service platforms needing high performance and computing power. The Java programming language will largely replace C, and yet another programming language will emerge to address technical challenges created by future algorithms, protocols, and
technologies.
With the huge increase in bandwidth availability, processing power, information, and interconnection provided by the network, a whole new class of applications and services that add value by transforming the information flowing through the network will become available. New signaling and routing protocols will be needed to address the high bandwidth, low delay, high interconnectivity, and high coordination demands placed on the network by these new technologies, applications, and services.
Many of the applications and services will also depend on the development of new algorithms and protocols for measuring and providing feedback on the perceived value and usefulness of information, distributed computing, parallel searching, and the coordination of resources to support these algorithms.
As these algorithms and protocols become widely implemented, the network will be able to independently organize and adapt itself in new and useful ways. In effect, the network will be able to apply something
akin to human judgment to organize, interpret, and disseminate information.
Service providers and carriers will differentiate into separate businesses, providing some combination of access to the network, transmission capacity, processing power in the form of service platforms, information repositories, and applications. Virtual networks will predominate as businesses and organizations make use of these new models to manage the storage, transmission, and processing of information to meet their needs.
At any instant in time, the service providers and carriers will have excess idle bandwidth and processing capacity. The sheer magnitude of this capacity will make it a commodity. Much like other commodities, a virtual real-time trading floor will develop for buying and selling transmission capacity (in the form of Mbps) and processing power (in the form of giga instructions/sec).
In 20 years, developing countries will provide the almost universal access currently enjoyed by developed countries.
Universal device mobility will be the norm. Universal access to people and information will have significant social and political consequences, generating privacy, security, economic, and national sovereignty policy issues, which individuals and governments will struggle to address.
The network is a work in progress. But one thing is certain - the future is likely to surprise us and be much more interesting than we can ever imagine today.
Jeff Lawrence is president and CEO of Trillium Digital
Systems, Inc in Los Angeles, CA. He received his BSEE from UCLA. Lawrence can be reached at j_lawrence@trillium.com.