An 8-issue superscalar core generally requires a 24-port RAM for the register file. The area and energy consumption of a multiported RAM increase in proportional to the square of the number of ports. A register cache can reduce the area and energy consumption of the register file. However, earlier register cache systems suffer from lower IPC caused by register cache misses. Thus, we proposed the Non-Latency-Oriented Register Cache System (NORCS) to solve the IPC problem with a modified pipeline. We evaluated NORCS mainly from the viewpoint of microarchitecture in the original article, and showed that NORCS maintains almost the same IPC as conventional register files. Researchers in NVIDIA adopted the same idea for their GPUs. However, the evaluation was not sufficient from the viewpoint of LSI design. In the original article, we used CACTI to evaluate the area and energy consumption. CACTI is a design space exploration tool for cache design, and adopts some rough approximations. Therefore, this paper shows design of NORCS with FreePDK45, an open source process design kit for 45nm technology. We performed manual layout of the memory cells and arrays of NORCS, and executed SPICE simulation with RC parasitics extracted from the layout. The results show that, from a full-port register file, an 8-entry NORCS achieves a 75.2% and 48.2% reduction in area and energy consumption, respectively. The results also include the latency which we did not present in our original article. The latencies of critical path is 307ps and 318ps for an 8-entry NORCS and a conventional multiported register file, respectively, when the same two cycles are allocated to register file read.

Wave absorber of rubber sheet containing natural rubber and EPDM is designed, fabricated and measured for improving ETC environment. As a result, proposed absorption material has fine weatherability and wave absorption satisfied with ETC standard can be realized theoretically before and after the weatherability test if the thickness of absorber is fabricated at the ranging from 2.26 mm to 2.52 mm. Moreover, absorber sheet sample based on theoretical values is fabricated and are measured. As a result, 20 dB or more is also confirmed at the incident angle ranging from 5 to 55 degrees experimentally. Therefore, the wave absorber with fine weatherability being satisfied with ETC standard can be realized.

The supertask approach was proposed by Moir and Ramamthy as a means of supporting non-migratory tasks in Pfair-scheduled systems. In this approach, tasks bound to the same processor are combined into a single server task, called a supertask, which is scheduled as an ordinary Pfair task. When a supertask is scheduled, one of its component tasks is selected for execution. In previous work, Holman et al. showed that component-task deadlines can be guaranteed by inflating each supertask's utilization. In addition, their experimental results showed that the required inflation factors should be small in practice. Consequently, the average inflation produced by their rules is much greater than that actually required by the supertasks. In this paper, we first propose a notion of Transient Behavior Prediction for supertasks, which predicts the latest possible finish time of subtasks that belong to supertasks. On the basis of the notion, we present an efficient schedulability algorithm for Pfair supertasks in which the deadlines of all component tasks can be guaranteed. In addition, we propose a task merging process which combines the unschedulable supertasks with some Pfair tasks; hence, a newly supertask can be scheduled in the system. Finally, we propose the new reweighting functions that can be used when the previous two methods fail. Our reweighting functions produce smaller inflation factor than the previous work does. To demonstrate the efficacy of the supertasking approach, we present the experimental evaluations of our algorithm, which decreases substantially a number of reweights and the size of inflation when there are many supertasks in the Pfair-scheduled systems.

We had previously proposed a fuzzy logic-based buffer management scheme (BMS) called fuzzy early detection (FED), which was designed to improve transmission control protocol (TCP) performance over the unspecified bit rate (UBR) service of asynchronous transfer mode (ATM) networks. Since a weakness in FED was discovered later, we present a refined version of it named FED+ here. Maintaining the design architecture and the algorithm of FED, FED+ further adopts a specific per virtual connection accounting algorithm to achieve its design goals. The effects of TCP implementation, TCP maximum segment size, switch buffer size and network propagation delay on FED+ performance are studied through simulation. Its performance is then compared with those of pure early packet discard (EPD), P-random early detection (P-RED) and FED. Our evaluations show that FED+ is superior to the others if the issues of efficiency, fairness, robustness, buffer requirement and the ease of tuning control parameters of a BMS are considered collectively.

A new fuzzy logic-based buffer management scheme (BMS) called fuzzy early detection (FED) is presented to improve transmission control protocol (TCP) performance over the unspecified bit rate (UBR) service of asynchronous transfer mode (ATM) networks. FED uses early congestion detection and selective packet discard (ECD/SPD) strategies to achieve its design goals, and adopts early packet discard (EPD) to prevent frequent buffer overflows. Two variants of FED are constructed for comparison purposes. FED performance is then evaluated against those of its variants, pure EPD and P-random early detection (P-RED) through various simulations. The effects of TCP implementation, TCP maximum segment size (MSS), network's propagation delay (PD) and switch buffer size (SBS) on their performances are studied. The overall result shows that FED is superior to the other BMSs when the criteria of efficiency, fairness, scalability and switch buffer requirement are considered collectively.

In order to ease the impact of the packet fragmentation problem and to avoid network congestion in TCP over UBR, packet discard schemes in ATM layer (such as PPD and EPD) have been proposed. These schemes drop packets before they reach their intended destinations if the network is congested and the packets are to be partially discarded. On the other hand, TCP also regulates data flow with its own flow control method. Due to restriction of data flow at the TCP layer, buffer space is not fully used in an ATM switch. In order to make use of more buffer resources, this paper generalizes the PPD and EDP schemes. From this generalization, an optimistic packet discard scheme named the "Probability-Based Delayed Packet Discard" (PDPD) scheme is proposed. Depending on a particular probability, this scheme sets a discard flag to delay actual discard operation. This paper presents the results of several simulated models to find out the potential of improvement of goodput by PDPD. The results of these simulations indicate that PDPD obtains higher goodput than ordinary schemes when the packet size is large and the input load is not light. This author concludes that a PDPD scheme should achieve effective goodput and link utilization while using more buffer resources effectively.

In this paper, the efficiency of transferring non-realtime data over Wireless ATM (WATM) networks is studied. Non-realtime services are sensitive to bit error as well as cell loss. The loss or error of a single cell due to congestion or a line error will result in the retransmission of the entire protocol data unit (PDU) by the end user in ATM networks. In WATM, cells are subject to the influences of noise. In this paper, we propose an adaptive cell checking controller (ACCC) for WATM networks to early find out error PDUs and to drop all remaining cells of these frames. The proposed ACCC only needs several bits overhead for each PDU of size of several Kilobytes. The removable percentage of an erroneous PDU by ACCC is analyzed. Simulation results show that compared with a conventional early packet drop mechanism, the proposed ACCC can achieve superior network utilization while keeping the minimum overhead in WATM networks.

Both available bit rate (ABR) service and unspecified bit rate (UBR) service with early packet discard (EPD) schemes have been considered for supporting data applications in ATM networks. Since transmission control protocol (TCP) is perhaps the most widely used transport layer protocol in existing data networks, the performance of TCP over ATM using ABR service and UBR service with EPD schemes is of great interest to ATM equipment vendors and service providers. In this paper, we present a simulation study of this interesting issue in a LAN environment using some benchmark network configurations proposed in the ATM Forum. Our simulation results show the following: (1) With UBR service and EPD schemes, TCP suffers significant performance degradation in terms of fairness and requires relatively large switch buffer even with a small number of active virtual connections over a LAN configuration, and (2) for the same set of network configurations and with ABR service using explicit rate feedback schemes, TCP achieves good performance in terms of fairness and link utilization, and requires relatively small switch buffer.

Rate-based congestion control is a promising scheme as data transfer service in ATM networks, and has been standardized in the ATM Forum. To migrate the existing upper layer protocols to ATM networks, however, further investigation is necessary. In particular, when ABR service class is applied to TCP (Transmission Control Protocol), duality of congestion control schemes in different protocol layers, i.e., conventional window-based congestion control in the Transport layer and ratebased congestion control in the ATM layer, may have a unexpected influence on performance. As an alternative approach for supporting TCP protocol, EPD (Early Packet Discard) has been recently proposed, which adds the function to the UBR (Unspecified Bit Rate) service. It does not have a "duality problem" since EPD only discards cells selectively to improve packet-level performance. In this paper, we exhibit performance of TCP protocol over ATM networks by using a simulation technique. We first compare rate-based control of ABR service and EPD applied to UBR service, and show that rate-control achieves better fairness and higher throughput in most circumstances. However, rate-based control requires careful tuning of control parameters to obtain its effectiveness and a duality problem leads to unexpected degradation of TCP-level performance. By the rate-based congestion control, temporal congestion at the switch is quickly relieved by the rate down of the source terminals. However, our simulation explores that if the parameter set of the rate-based congestion control is not appropriately used, the congestion is also recognized at TCP due to packet drops and TCP unnecessarily throttles its window size. To avoid this sort of the problem, we develop the appropriate parameter set suitable to TCP on ABR service, and point out that some modification of TCP may be necessary for further performance improvement.