Large Scale Networks and Future Network Architectures -session summary. Jim Kurose Department of Computer Science University of Massachusetts Amherst MA http: //www. cs. umass. edu/~kurose
Architecture: What did we hear? q focus on control v v v signaling complexity data plane interaction routing cross layer q Qo. S/multimedia v various frameworks advocated: no consensus v protocol need q Role-based architecture What didn’t we hear? q optical nets v o-core, e-edge v grid/e-science v GMPLS, circuits ondemand q management, AAA q security q Reliability q Economic incentives and tradeoffs
Architecture: comments heard q system complexity: component versus system complexity v bigger picture q new inter/intra domain algorithms q beyond Internet v v v Internet just one network instance But it is “the” network of today, and needs our research efforts Different size/type of adhoc a place for network architects? Internet protocols versus pverall architecture q sensor nets: v clean slate? v lessons from the past? v sensor != wireless q need for new transport protocols v ad hoc, sensor, MM,
Education: comments heard q what should we be teaching our grad students v FYI: SIGCOMM education workshop q internet-inculcation of students q learn from the past v SS 7, ATM, SNA, DECNET, ….