Fuzzing is an effective technique to discover software vulnerabilities by automatically generating large volumes of inputs that may trigger crashes. However, post-fuzzing root cause analysis (RCA) to identify the true problems of the crashes requires substantial manual efforts, leaving many crashes discovered through fuzzing unresolved. Previous studies based on statistical approaches assist human analysts by suggesting candidate root causes inferred from large volumes of fuzzing results, but they often fail to reveal the causal relationship between the root cause and the crash site, as the two are typically distant in the code and connected only indirectly. We propose Sagitta, an assistive tool to facilitate post-fuzzing RCA by highlighting highly probable root causes based on differences in data flow between crashing and non-crashing inputs. Sagitta uses dynamic taint analysis on a crashing input and its pre-mutation parent to identify diverging points in data flow as well as control flow, and visualizes the results as localization graphs enriched with lexical information to aid analyst comprehension. We evaluated Sagitta on 14 known vulnerabilities in large-scale software and found that it captures causal relationships with only 1 to 10 branches across five bug categories. The results demonstrate that visualizing data flow differences with Sagitta's localization graphs significantly reduces manual effort in RCA.

Mobile broadband networks are becoming increasingly popular, and they support many different IP based services. The Transmission Control Protocol (TCP) is a widely used protocol in the Internet. In this paper, we analyze the performance of the existing loss-based and delay-based congestion control algorithms for TCP on real mobile networks. The impact the loss-based algorithm has in terms of network latency is extremely heavy and degrades the communication quality of all flows which share the same bottleneck link. On the other hand, although delay-based TCP is able to maintain latency at a low level, its goodput is very small because the RTT increases independently of network congestion. We propose a new delay-based congestion control algorithm. To enable this algorithm to work well on mobile networks, we use a variable threshold which includes the delay caused by the data link layer. Experiments on emulated network and real network show that the algorithm can maintain the queue size and transmission delay at low levels without degrading goodput.

This paper describes the recent progress of research and development on transmission technologies for Super Hi-Vision (SHV), which is the form of ultrahigh-definition television (UHDTV) with the highest resolution class. The paper covers the transmission systems for SHV program production and satellite broadcasting. For short-haul transmission, an optical digital interface for video equipment was developed. It wavelength-multiplexes eight 10-Gb/s optical signals. It was experimentally confirmed that the transmission distance was at least 2 km, and that a series connection of the interfaces did not severely deteriorate jitter performance. We are also developing a 120-GHz-band wireless contribution short-haul link consisting of three transmitter/receiver pairs. All three pairs transmit and receive a 10.3-Gb/s signal at the same frequency. Experiments showed that transmission with little crosstalk is possible over a distance of 1.25 km when the receiver spacing is 8 m. A long-haul optical contribution link that we developed can transmit a 24-Gb/s SHV signal over a distance of 260 km by using optical amplifiers and no electronic repeaters. An IP network would be a good choice for long-distance links such as an international live contribution link. Several live IP transmissions of compressed video were conducted. Satellite relay transmission and broadcasting technology is also being developed. In particular, we have developed a QPSK/8PSK wide-band modem that uses LDPC and BCH codes. The modem could transmit 251-Mb/s (QPSK) and 677-Mb/s (8PSK) signals and was used in a live SHV satellite-transmission experiment.

IP multicast technology enables simultaneous delivery of multimedia content to numerous users. However, the quality of experience (QoE) is subject to degradation due to packet loss because quality of service (QoS) is not guaranteed in best-effort IP networks. To maintain good QoE, streaming applications typically use an application-layer FEC (forward error correction) technique. Conventional FEC techniques always transfer redundant packets, but these redundant packets are only used when packet loss has occurred. This causes inefficiency in networks where packets are not often lost. To avoid filling network paths with unnecessary redundant packets, we developed a new application-layer packet loss recovery mechanism based on layered multicast using FEC and ARQ. This mechanism divides redundant packets into fundamental and supplementary redundant packet groups and sends them in separate multicast sessions. The supplementary redundant packets are received only by receivers participating in the multicast session of them or requesting retransmission. This mechanism realizes efficient reliable multicast for HDTV IP broadcasting by reducing unnecessary redundant packets and preventing an explosion of requests. We tested the effectiveness of our mechanism in streaming HDTV experiments over a commercial best-effort IP network. The results of experiment show that the proposed mechanism avoids transferring wasteful, redundant packets and efficiently maintains reliability with preventing an explosion of requests.

Multicast packets are lost when collisions occur on IEEE 802.11 wireless LANs with DCF because the IEEE 802.11 MAC protocol does not use a packet loss recovery mechanism for multicast packets. Therefore, an application that uses wireless multicast should be provided with a recovery mechanism. This paper studies the multicast packet loss due to collision with TCP packets on IEEE 802.11b DCF infrastructure wireless LANs. The results of analysis and simulation show that the multicast packet loss ratio has an upper limit, and that FEC is effective for recovering the multicast packet losses.