This paper presents a network surveillance technique for detecting malicious activities. Based on the hypothesis that unusual conducts like system exploitation will trigger an abnormal network pattern, we try to detect this anomalous network traffic pattern as a sign of malicious, or at least suspicious activities. Capturing and analyzing of a network traffic pattern is implemented with a concept of port profiling, where measures representing various characteristics of connections are monitored and recorded for each port. Though the generation of the port profiles requires the minimum calculation and memory, they exhibit high stability and robustness. Each port profile retains the patterns of the corresponding connections precisely, even if the connections demonstrate multi-modal characteristics. By comparing the pattern exhibited by live traffic with the expected behavior recorded in the profile, intrusive activities like compromising backdoors or invoking trojan programs are successfully detected.

The main issues of the next generation Internet are high performance multicast, security and QoS (Quality of Service). The network layer multicast is accepted as a powerful solution to reduce communication overhead in networks as well as hosts. Although a lot of multicast routing protocols have been developed and have worked for the current Internet, they have drawbacks in scalability, reliability and network load balancing. In this paper, we propose an enhanced multicast routing protocol called DCBT (Decentralized Core based Tree), which includes some features of CBT (Core Based Tree), PIM-SM (Protocol Independent Multicast-Sparse Mode) and BGMP (Border Gateway Multicast Protocol). Fundamentally DCBT differs from those in that it covers not only intra-region but also inter-region multicast. RNs (Regional networks) have their intra-region shared trees, which are interconnected by an inter-region shared tree for a multicast group. New two concepts are introduced: the one is RBTS (Region Based Tree Switching) to achieve network load balancing and the other is to redefine TOS (Type Of Service) byte to make special multicast routing decision. In addition, the effect of core location is given by changing its position within an RN. Finally we provide the performance evaluation of DCBT by OPNET network simulator over a real scale network model.

TCP Redirection (TCP-R) is an extension of TCP, which maintains active TCP connections when the disconnection occurs due to IP address changes. In mobile computing environments, IP addresses may be changed frequently as the host moves across subnets or changes network devices. However, it is difficult for most stateful network applications to work continuously in that situation. There is much research to support such dynamics in the network layer, but these attempts in the IP layer tend to incur some complexity. This paper intends to investigate the end-to-end mobility support in the transport layer. We developed a simple redirection mechanism for TCP, which enables us to keep our working activities continuous without any intermediate agents for IP packet forwarding. We also show that TCP-R enables the intentional replacement of IP addresses, which is useful for certain network services. We implemented and evaluated a prototype of TCP-R. The measured performance indicates that TCP-R can realize continuous operation with minimal overhead and complexity.

In deep submicron CMOS and BICMOS technologies, antenna effects affect floating gate charge of usual floating gate test structures, dedicated to capacitor matching measurement. In this paper a new pseudo-floating gate test structure is designed. After test structure and modeling presentation, testing method and results are given for several capacitor layouts (poly-poly and metal-metal).

Multicasting has become an important part of current networking technologies. To accommodate the multicasting traffic, switches in ATM (Asynchronous Transfer Mode) networks need to be able to support multicasting. In multicast ATM switches, usually there are many cells contending for the same output port resources at the same time. This may lead to serious performance degradation. In the past, many special output port components are proposed to solve this problem. However, most of these approaches alleviate the output port overflowing problem without considering the output port congestion or the internal routing blocking problem. Especially when the switch fabric is based on the banyan-like network, these two problems will seriously degrade the switch performance. In this paper, we propose a pre-scheduling mechanism to precede the multicast switch. This mechanism will completely eliminate the problems mentioned above. Two architectures are designed to implement this mechanism. Analysis and simulations show that our design performs quite well.

This paper proposes a traffic-adaptive dynamic routing method, which we have named RAG, for connectionless packet networks. Conventional traffic control methods discard the packets which cause congestion. Furthermore, conventional routing methods propagate control messages all over the network for gathering global topology information, and this causes more congestion. In contrast, RAG estimates traffic conditions all over a network without any communication between nodes and makes the best use of free links so that packets make detours to avoid congestive sites. RAG adopts distributed control based on game theory (non-communication, non-zero-sum, two-person). With RAG, nodes play a packet-forwarding game without any communication with each other, and each node controls ordering and routing of the forwarding packets based on the node's individual payoff table which is dynamically reconstructed by observation of surrounding nodes. Nodes cooperate with each other, except for punishment for disloyalty. Repetition of these local operations in nodes aims at the emergence of the gradual network-global traffic balancing. The results of experiments in comparison with the conventional shortest path first (SPF) routing method show that the throughput is about 1.58 times higher with the new method.

A novel method to enhance the practical security of interferometric quantum cryptography is proposed, giving the protocol and detailed constructions including a controlled spontaneous photon emitter, a superradiance amplifier, beam splitters, phase shifters, and a pair of Mach-Zehnder interferometers. The intrinsic uncertainty due to the random phase selection out of three, leads to the detection of eavesdropping. The physical uncertainty of the controlled spontaneous emission of coherent photons also adds temporal equivocation to confuse eavesdroppers.

A new simple method for extracting the capacitance coupling coefficients of sub-0.5-µm flash memory cells is proposed. Different from the previously proposed methods, this method is not affected by a dopant profile of source region because a band-to-band tunneling current from the interface between the drain and the substrate is probed. Use of a reference device eliminates the necessity to make assumptions concerning the electron transport mechanism. Comparison with the other methods shows that the proposed method is simple and accurate.

We have investigated the operation of a reflection type magnetostatic wave signal-to-noise enhancer in detail. It has good enhancement characteristics, low insertion loss, and low operating power. It is also composed of a transducer using a ceramic substrate having a high dielectric constant and an LaGa-YIG film with low saturation magnetization to enable direct operation in the 400-MHz band (the IF band of current DBS receivers). Enhancement of 8 dB was achieved over a 40-MHz bandwidth. Although its operating frequency range depends critically on device temperature, we can compensate for the temperature dependence by adjusting the bias magnetic field. Experiments showed that the enhancer improved the received carrier-to-noise ratio by 2 to 3 dB, providing good noise reduction in DBS reception.

It is confirmed by simulation that SOI-DRAMs can be operated at high speed by using the floating body structures. Several floating body effects are analyzed. It is described that the dynamic retention characteristics are not dominated by capacitive coupling and hole redistribution. And it is described that those characteristics are determined by the leakage current in the small pn-junction region of the floating body. Reducing pn junction leakage current is important for realizing a long retention time. If the pn junction leakage is suppressed to 10-18 A/µm, a dynamic retention time of 520 sec at a VBSG of 0.5 V can be achieved at 27. On those conditions, the refresh current of SOI-DRAMs is reduced by 54% compared with bulk-Si DRAMs.

Computational sensor (smart sensor, vision chip in other words) is a very small integrated system, in which processing and sensing are unified on a single VLSI chip. It is designed for a specific targeted application. Research activities of computational sensor are described in this paper. There have been quite a few proposals and implementations in computational sensors. Firstly, their approaches are summarized from several points of view, such as advantage vs. disadvantage, neural vs. functional, architecture, analog vs. digital, local vs. global processing, imaging vs. processing, new processing paradigms. Then, several examples are introduced which are spatial processings, temporal processings, A/D conversions, programmable computational sensors. Finally, the paper is concluded.

In this paper, we propose a novel fault-tolerant multicast algorithm for n-dimensional wormhole routed hypercubes. The multicast algorithm will remain functional if the number of faulty nodes in an n-dimensional hypercube is less than n. Multicast is the delivery of the same message from one source node to an arbitrary number of destination nodes. Recently, wormhole routing has become one of the most popular switching techniques in new generation multicomputers. Previous researches have focused on fault-tolerant one-to-one routing algorithms for n-dimensional meshes. However, little research has been done on fault-tolerant one-to-many (multicast) routing algorithms due to the difficulty in achieving deadlock-free routing on faulty networks. We will develop such an algorithm for faulty hypercubes. Our approach is not based on adding physical or virtual channels to the network topology. Instead, we integrate several techniques such as partitioning of nodes, partitioning of channels, node label assignments, and dual-path multicast to achieve fault tolerance. Both theoretical analysis and simulation are performed to demonstrate the effectiveness of the proposed algorithm.

The tera-bit order capacity of ultrahigh-speed and wide-band networks will become necessary to provide highly advanced multimedia services. In conventional networks, electronic circuits limit the speed capability of the networks. Consequently, all-optical networks are essential to realize ultrahigh-speed and wide-band communications. In this paper, we propose the configuration of an all-optical code division multiplexing (CDM) switching network based on self-routing principles and the structure of a nonlinear all-optical switching device as one of the key components for the network. We show that the required performances of the optical devices used in the CDM switching fabric are lower than those used in the TDM and illustrate the basic transmission characteristics of the switching device utilizing FD-BPM. To evaluate the multiplexing performance, we demonstrate the maximum number of channels under an error-free condition and the BER characteristics when the Gold sequence is applied as one of the CDM code sets, and show that the network of the sub-tera-bit order capacity is realizable by adopting TDM, WDM and CDM technologies. We also illustrate the packet assembly method suitable for self-routing transmissions and one of network architectures where the proposed switching fabric can be exploited.

Given two terms and their rewriting rules, an unreachability problem proves the non-existence of a reduction sequence from one term to another. This paper formalizes a method for solving unreachability problems by abstraction; i. e. , reducing an original concrete unreachability problem to a simpler abstract unreachability problem to prove the unreachability of the original concrete problem if the abstract unreachability is proved. The class of rewriting systems discussed in this paper is called β rewriting systems. The class of β rewriting systems includes very important systems such as semi-Thue systems and Petri Nets. Abstract rewriting systems are also a subclass of β rewriting systems. A β rewriting system is defined on axiomatically formulated base structures, called β structures, which are used to formalize the concepts of "contexts" and "replacement," which are common to many rewritten objects. Each domain underlying semi-Thue systems, Petri Nets, and other rewriting systems are formalized by a β structure. A concept of homomorphisms from a β structure (a concrete domain) to a β structure (an abstract domain) is introduced. A homomorphism theorem (Theorem1)is established for β rewriting systems, which states that concrete reachability implies abstract reachability. An unreachability theorem (Corollary1) is also proved for β rewriting systems. It is the contraposition of the homomorphism theorem, i. e. , it says that abstract unreachability implies concrete unreachability. The unreachability theorem is used to solve two unreachability problems: a coffee bean puzzle and a checker board puzzle.

The application of photonic technologies to packet switching offers the potential of very large switch capacity in the terabit per second range. The merging of packet switching with photonic technologies opens up the possibility of packet switching in transparent photonic media, in which packets remain in optical form without undergoing optoelectronic conversion. This paper reviews recent work on photonic packet switching. Different approaches to photonic packet switching and key design issues are discussed.

Design and evaluation of a high-performance switch architecture for free-space photonic switching systems is described. The switch is constructed of 22 switching elements and employs special multistage interconnection patterns. The connection setup algorithm and the control procedure at the switching elements are based on a "rerouting algorithm." Performance analysis shows that the blocking probability of the switch is easily controlled by increasing the number of switching stages. Example implementations of this switch are shown in which birefringent plates, polarization controllers, etc. are used.

Multi-input floating gate differential amplifier (FGDA) is proposed which can perform any convolution operation with differential structure and feedback loop. All operations are in the voltage mode. Only one terminal is required for the negative feedback which can suppress distortions due to mismatches of active elements. Possible applications include intelligent image sensor, where fully parallel DCT operation can be performed. A prototype chip is fabricated which is functional. A preliminary test result is reported.

Recently, mobile computing has received much attention from database community. Sharing information among mobile users is one of the most challenging issues in mobile computing due to user mobility. Replication is a promising technique to this issue. However, adopting replication into mobile computing is a non-trivial task, since we are still facing other problems such as the lack in disk capacity and wireless network bandwidth used by mobile users. We have proposed a dynamic replica allocation strategy called User Majority Replica Allocation (UMRA) that is well suited to the modern architecture of mobile computing environment while avoiding such problems mentioned above. In this paper, we propose two relocation decision policies for UMRA and we provide a cost analysis for them. We also provide a cost analysis for another replica allocation strategy called Static Replica Allocation (SRA) for a comparison purpose.

Operation of fiber-grating semiconductor laser (FGL) has been stabilized by using the semiconductor optical amplifier which has a simple slant-waveguide structure. The emission wavelength, which depends on a temperature, shows hysteresis. Employing the directly modulated FGL at 2.5 Gb/s, transmission over 400 km in standard optical fiber has been successfully achieved.

The tera-bit order capacity of ultrahigh-speed and wide-band networks will become necessary to provide highly advanced multimedia services. In conventional networks, electronic circuits limit the speed capability of the networks. Consequently, all-optical networks are essential to realize ultrahigh-speed and wide-band communications. In this paper, we propose the configuration of an all-optical code division multiplexing (CDM) switching network based on self-routing principles and the structure of a nonlinear all-optical switching device as one of the key components for the network. We show that the required performances of the optical devices used in the CDM switching fabric are lower than those used in the TDM and illustrate the basic transmission characteristics of the switching device utilizing FD-BPM. To evaluate the multiplexing performance, we demonstrate the maximum number of channels under an error-free condition and the BER characteristics when the Gold sequence is applied as one of the CDM code sets, and show that the network of the sub-tera-bit order capacity is realizable by adopting TDM, WDM and CDM technologies. We also illustrate the packet assembly method suitable for self-routing transmissions and one of network architectures where the proposed switching fabric can be exploited.