Secure Multicast Services and Applications

点击翻译按钮获取中文摘要

This is a collaborative research proposal being submitted by the University of Texas at Austin(PI: Simon S. Lam) and the University of Kentucky (PI: Kenneth Calvert).Most Internet applications are based upon the client-server paradigm and make use of unicastdelivery of data sent as independent packets. The technical issues of securing unicast communi-cations for client-server computing are fairly well understood. However, many emerging Internetapplications are based upon a group communications model. In particular, they require packetdelivery from one or more authorized senders to a large number of authorized receivers. Further-more, the data of many of these applications should be viewed as flows rather than as independentpackets. Conceptually, since every point-to-multipoint communication can be represented as a setof point-to-point communications, and every flow can be treated as a sequence of individual pack-ets, the current technology base for securing unicast packets can be extended in a straightforwardmanner to securing applications that employ flows and multicasts. Such extensions, however, wouldbe highly inefficient. They will not be applicable to high-speed transmission and delay-sensitiveows, nor scalable to large groups.The objectives of this proposal are to investigate architectures and protocols for securing mul-ticast and flow applications with emphasis on efficiency, latency, and scalability. The researchers have obtained some preliminary results on scalable group key management and efficient digital signatures. Theseare essential ingredients for designing and implementing secure multicast services. However, to de-sign a scalable service for secure multicast on the Internet, even with just best-effort delivery, severalresearch issues remain, namely, performance comparison of alternative architectures, efficient andreliable delivery of rekey messages (and their impact on a wide area network), and deployment oftrusted components and active agents.The researchers next propose to investigate the more difficult problem of reliable secure multicast. For theInternet, this problem has been largely unexplored. Initially, the researchers plan to approach this problemfrom two directions, namely, (i) how to make reliable multicast secure, and (ii) how to make securemulticast reliable. In addressing this problem, their concerns are, first, correctness, and second,performance issues such as latency and scalability.There are many applications that can benefit from techniques and services for securing flowsand multicasts, e.g., teleconference, pay per view, informaton services, virtual private networks,collaborative work, software distribution, etc. The researchers propose to experiment with some of these applications. In particular, they propose to investigate a less obvious application, namely, active networks,which raises some interesting research issues. Securing active networks is an important problemwith many different concerns, some of which are being addressed by other researchers. In the pro-posed project, the researchers are primarily concerned with characterizing the authenticity and confidentiality requirements of mobile code distribution in active networks and ensuring their satisfaction using efficient methods.