In this paper we study the recording and watching patterns of DVR users using real traces. Many DVR users complain about running out of space. Thus, we propose the idea of adding extra storage on the network. Storing content in the network would create a new challenge, as most of users use their DVR during prime-time resulting in congesting the already strained network with unicast streams. We develop statistical models that learn the behavior of users and are most likely to make watched programs readily available for users each day. We develop a simple caching technique that captures the dominant factors of user behavior. Our verification, using real traces, shows that this technique performs as efficiently as more advanced statistical models, while requiring only a small state to be maintained.

In science and engineering courses, students are often presented a situation for which they are asked to identify the relevant principles and to instantiate them as a set of equations. For an ITS to determine the correctness and rel- evance of the student's answer and generate effective feed- back, it must map the student variables and equations onto the physical properties and concepts that are relevant to the situation. The space of possible mappings of variables and equations is extremely large. Domain independent tech- niques by themselves are unable to overcome the complexity hurdles. This paper describes how an ITS can use constraint propagation and algebraic techniques combined with do- main and problem-specific knowledge to solve the mapping problem with systems of algebraic equations. The tech- niques described in this paper have been implemented in the PHYSICS TUTOR tutoring system and evaluated on three data sets that contain submissions from students in several introductory Physics courses.

IP networks may soon become a delivery mechanism for broadcast television content. Multicast can reduce the steady state bandwidth demand on network links from one stream per viewer to one stream per watched program. However, channel surfing at commercial breaks can periodically increase the bandwidth demand. In the channel change mechanism we study, surfers leave multicast groups and receive unicast streams at higher than usual bandwidth. This paper builds a mathematical model to determine the net bandwidth demand of multicast and surfing during commercial breaks. In one example, we find that the peak demand during a commercial break is twice the steady state multicast demand.

This paper analyzes implementations of automatic code gapping (ACG) in intelligent network service control points (SCPs). The objective is to identify ACG controls that provide high call throughput while protecting the network from overloads. To achieve this we emphasize the need for careful engineering of vendors' control implementations (for example, the proper choice and use of internal SCP measurements) within the constraints of present ACG requirements. We use simulation and find that algorithms that use a combination of processor utilization measurements and measurements of excessive response times can yield reasonable call throughput. Shortening the measurement intervals to speed up the rate at which the SCP can change the rate at which traffic arrives to it further improves the call throughput. Finally, making the sizes of the rate changes smaller refines the granularity of the control and leads to better results.

The ITU-T (International Telecommunications Union, Telecommunications Standardization Sector, formerly known as CCITT), in its specification of the SS7 protocol, describes a number of alternatives for congestion control. In this paper we provide a direct comparison of the effectiveness of some of these methods in dealing with a mass call-in situation. We find that the international option provides better call throughput and is more robust to threshold settings and delays in feeding back congestion information than the national option with congestion priorities and a combination of the two options. The results can provide guidance to those attempting to define application level overload procedures, to those studying internetworking issues for connection between countries employing different SS7 congestion control options, and to those preparing to formulate B-ISDN overload strategies.>

The paper analyzes the performance of two different types of service control point (SCP) overload control algorithms, table-driven controls and an adaptive control. As performance measures, the author uses aggregate transaction throughput and fairness (whether the control algorithm allocates SCP capacity to individual traffic sources as desired). It is found that, even when parameters are well-matched to source characteristics, table-driven controls do not perform as well as the adaptive control. In addition, the adaptive control is more robust to traffic patterns and SCP internal operation than the table-driven controls.>

This paper develops a queueing model of a buffer that collects cells for reassembly into packets for a protocol layer above the asynchronous transfer mode (ATM) layer. Whenever the buffer fills with all packets incomplete, a packet must be sacrificed to make room for others. The queueing model estimates the equilibrium fraction of packets sacrificed under one algorithm for selecting the packet to be sacrificed. The paper also uses simulation to compare three sacrifice algorithms. The model's predicted packet loss probabilities bound from above the loss probabilities in the simulations of the different algorithms. Applications to sizing the buffer for a prescribed loss probability are given.

This paper studies the performance of the common channel signaling link congestion control procedure in the presence of network latency (the delay between the onset of congestion and the reaction of traffic sources to congestion information). The context of the study is a focused overload on a signaling point. The results show that the procedure tends to synchronize the intervals of time when traffic sources may or may not transmit and it tends to overcontrol traffic. Consequently, call (not just message) throughput suffers. The paper proposes and analyzes several partial remedies that are easy to implement. In particular, it recommends altering timer values and queue congestion thresholds and finds that the call throughput improves dramatically.>

It is desirable to interconnect different computer hosts and local area networks (LANs) through the asynchronous transfer mode (ATM) network via broadband terminal adaptors (BTAs). The BTA must have a sufficiently large buffer, called a virtual channel queue (VCQ), to temporarily store multiple, partially received packets from different virtual channels. The buffer requirement of a shared-memory VCQ is studied for different packet loss probabilities and virtual channel numbers. Two different architectures for implementing the shared-memory VCQ are compared. The second architecture with multiple linked queues in the shared-memory requires less memory and has better scalability to accommodate a large number of virtual channels and is adopted in the analysis. Several possible error conditions, such as shared-memory overflow, the received packet exceeding its maximum length, and the corruption of the pointer in the logical queue, are discussed. Corresponding solutions are proposed in the VCQ designs.>

The authors develop a queuing model of a buffer that collects cells for reassembly into packets for a protocol layer above the asynchronous transfer mode (ATM) layer. Whenever the buffer fills with all packets incomplete, a packet must be sacrificed to make room for others. The queuing model estimates the equilibrium fraction of packets sacrificed. Applications to sizing the buffer for a prescribed loss probability are given.>