Zheng Shaoren

Analysis and evaluation of the TCP/IP protocol stack of LINUX

With the rapid progress of high-speed network technology, the communication speed between hosts will reach Gbit/s. So it is important to investigate the traditional TCP/IP protocol stack to find out whether it is suitable for high-speed communication and can provide QoS or not. Based on this consideration, the LINUX TCP/IP stacks structure, the buffer management, the procedures for sending and receiving data, and the interaction with the OS are analyzed in this paper. A probing node method is proposed to investigate the internal structure of the stack. With this method, the throughput of the stack, the time consuming of each layer and the queuing of the IP layer at high speed are tested. Our experiments show that the throughput of the TCP/IP stack of LINUX can reach 360 Mbit/s at the sending side and 150 Mbit/s when the local host device is used for bi-directional operation, the protocol stack can he improved to support higher speed; the percentages of the various costs of the TCP/IP protocol suite have changed greatly, zero-copy and zero-scan techniques cannot improve the performance dramatically; there is no queuing at the receiving side of the IP layer, so FCFS (first come first served scheduler) is enough, but a more sophisticated packet scheduler such as WFQ (weighted fair queuing) is needed at the sending layer of the IP layer to provide QoS in the future.

A novel method for label switching

A novel model is introduced to implement label switching. It is a model that implements both control-driven and data-driven label switching approaches. We call this novel model the PLS (partial label switching) model. Compared to models such as MPLS, the PLS model may save approximately 75%-88% of VC space, with a little sacrifice (5% to 17%) to the packet-forwarding capacity. At the same time, using the same label space, the PLS model costs 35% less packet-forwarding capacity than the IP switching model. From simulation results we can conclude that the router will reach higher performance if it uses the PLS model.

Dynamic compression, and storage of voice information in the distributed voice information database

The management and maintenance the voice information is an important subject of the voice mailbox system. A dynamic compression algorithm is presented. The algorithm compresses the voice information by two steps. In the first step, a zero detection technique is used to delete the blank places in the voice signal. In the second step, optimum prediction and adaptive quantization techniques are used to compress the sampling rate of the voice signal. A new voice mailbox system architecture is discussed, the performance of which is analysed and the results given. The database which stores the voice signal in our implementation is a distributed LAN, which is composed of independent microcomputers.

Weighted max-min fair scheduling in input-queued crossbar switches

The input-queued switching architecture, because of its scalability, has become attractive for implementing high-speed switches and routers. For an input-queued switch, it is challenging to provide a scheduling scheme that is both highly efficient and fair in bandwidth allocation. Existing schedulers attempt to improve throughput by scheduling cells free of contention for transmission, at the expense of delay and fairness. In this paper, we proposed a weighted max-min fair scheduling (WMFS) scheme for input-queued switches that provides weighted max-min fair bandwidth allocation distribution among the flows. Results from simulation show that the scheme is able to achieve asymptotically 100% throughput and low cell latency under uniform traffic, and, provide approximately weighted max-min fair bandwidth allocation and normalized fair service guarantees in the presence of misbehaving traffic.

The analysis of voice loss ratio and data delay in integrated service PRMA

The stationary status performance of FDD integrated service PRMA is analyzed based on the simplified six-status Markovian model; the formulas of voice packet loss ratio and data packet delay are also obtained. Through numerical calculation and simulation, the impact of permit probability, channel slot number, the maximum voice delay to system performance is also studied. From the two aspects of voice loss ratio and data delay, it can be seen that the system gain is about 0.75. In order to guarantee the quality of voice and data, the number of data stations should be controlled to be below that of voice stations.

The implementation of ISDN access signalling

It is an overwhelming trend that the present telecommunications network is evolving towards the integrated services digital network. The user requirement for telecommunication services is expanding. The old user-network signaling system in the public switched telephone network (PSTN) can no longer meet the needs of various new services, therefore the ISDN access signaling system must be implemented. This paper describes the implementation of the ISDN primary-rate access signaling in the HJD-1000A stored program controlled switches.

The implementation of SS7 in JSU-16000

An important aspect in the development of communication networks is the development of signalling technology. SS7 is a common channel signalling system recommended by CCITT, which is adapted to the development of ISDN and IN. This paper presents a general scheme to implement SS7 in JSU-16000, a fully distributed control digital switching system, that is developed by ourselves.