In video conference systems over the Internet, audio and video data are often lost due to UDP packet losses, resulting in degradation of assurance. In this paper we describe a high-assurance video conference system applying the following two techniques: (1) packet loss recovery using convolutional codes, which improves the assurance of communication; and (2) Xcast, a multicast scheme that is designed for relatively small groups, reducing the bandwidth required for a multi-point conference. We added these functions to a GateKeeper (GK), a device used in conventional conference systems. Encoding/decoding and Xcast routing were then implemented as the upper layer for the UDP. We examined the functions of the system over the Internet in a multi-point conference between three sites around Tokyo, as well as a conference between Tokyo and Korea. We also investigated the effectiveness of the proposed system in experiments using an Internet simulator. Experimental results showed that the quality of received picture was improved in comparison with the case where no encoding schemes were applied.

In numerical simulations using massively parallel computers like GPGPU (General-Purpose computing on Graphics Processing Units), we often need to transfer computational results from external devices such as GPUs to the main memory or secondary storage of the host machine. Since size of the computation results is sometimes unacceptably large to hold them, it is desired that the data is compressed and stored. In addition, considering overheads for transferring data between the devices and host memories, it is preferable that the data is compressed in a part of parallel computation performed on the devices. Traditional compression methods for floating-point numbers do not always show good parallelism. In this paper, we propose a new compression method for massively-parallel simulations running on GPUs, in which we combine a few successive floating-point numbers and interleave them to improve compression efficiency. We also present numerical examples of compression ratio and throughput obtained from experimental implementations of the proposed method runnig on CPUs and GPUs.

A Monolithic amplifier IC with small phase deviation applying parallel feed back technique was developed by using high speed Si bipolar process technology. This IC achieves a phase deviation of 1.8, an S21 gain of 46 dB and a limiting output of 7.8 dBm at 1 GHz.

This paper discusses distributed checkpointing with logging for practical applications running with limited resources. We present a discrete time model evaluating the total expected overhead per event where the number of available checkpoints that each process can hold is finite. The rollback distance is also bound to some finite interval in many actual applications. Therefore, the recovery overhead for the checkpointing scheme is described by using a truncated geometric distribution as the rollback distance distribution. Although it is difficult to analytically derive the optimal checkpoint interval, which minimizes the total expected overhead, substituting other simple probabilistic distributions instead of the truncated geometric distribution enables us to do this explicitly. Numerical examples obtained through simulations are presented to show that we can achieve almost minimized total overhead by using the new models and analyses.

An approach is proposed for constructing a dependable server cluster composed only of server nodes with all nodes running the same algorithm. The cluster propagates an IP multicast address as the server address, and clients multicast requests to the cluster. A local proxy running on each client machine enables conventional client software designed for unicasting to communicate with the cluster without having to be modified. Evaluation of a prototype system providing domain name service showed that a cluster using this technique has high dependability with acceptable performance degradation.

This paper discusses hybrid state saving for applications in which processes should create checkpoints at constant intervals and can hold a finite number of checkpoints. We propose a reclamation technique for checkpoint space, that provides effective checkpoint time arrangements for a rollback distance distribution. Numerical examples show that when we cannot use the optimal checkpoint interval due to the system requirements, the proposed technique can achieve lower expected overhead compared to the conventional technique without considering the form of the rollback distance distribution.