ISSN 
2004 Vol. 33, No. 6
2004, 33(6): 627-632.
Abstract:
Optical Code Division Multiple Access (OCDMA) Communications Technology is introduced in detail. The research results of the achievement of Wavelenght Division Multiple Access/Optical Code Division Multiple Access (WDM/OCDMA) hybrid networks, construction for OOCs, multiple access interference(MAI) and bipolar OCDMA system studied. Finally, some predominances and key techniques are proposed.
Optical Code Division Multiple Access (OCDMA) Communications Technology is introduced in detail. The research results of the achievement of Wavelenght Division Multiple Access/Optical Code Division Multiple Access (WDM/OCDMA) hybrid networks, construction for OOCs, multiple access interference(MAI) and bipolar OCDMA system studied. Finally, some predominances and key techniques are proposed.
2004, 33(6): 633-638.
Abstract:
Optical fiber access network is developed rapidly in recent ten years and the new technology solves the last 1 km transmission from optical transmission network to customers. The broadband optical fiber access network is developed from analog to digital system, the representative of analog signal optical fiber access network is HFC (hybrid fiber coaxial) network, and the representative of digital signal optical fiber access network is PON(passive optical network). HFC is a developing technology from the middle 90's last century and has been used extensively. PON is a new technology risen in recent years. PONs includes APON (ATM PON), EPON (Ethernet PON) and GPON (Gigabit PON). The key techniques and features of these access networks are analyzed and discussed in detail in this paper.
Optical fiber access network is developed rapidly in recent ten years and the new technology solves the last 1 km transmission from optical transmission network to customers. The broadband optical fiber access network is developed from analog to digital system, the representative of analog signal optical fiber access network is HFC (hybrid fiber coaxial) network, and the representative of digital signal optical fiber access network is PON(passive optical network). HFC is a developing technology from the middle 90's last century and has been used extensively. PON is a new technology risen in recent years. PONs includes APON (ATM PON), EPON (Ethernet PON) and GPON (Gigabit PON). The key techniques and features of these access networks are analyzed and discussed in detail in this paper.
2004, 33(6): 639-642.
Abstract:
In this paper, A new low-loss wide-angle Y-Junction with graded-index optical waveguides, which can change the optical propagation direction, is introduced. An application of the Y-junction structure to 1×4 optical splitter and Mach-Zehnder modulator is analyzed. The two kinds of device performance are examined numerically by the finite difference beam propagation method (FDBPM). The simulated results show that the 1×4 optical splitter has very low loss of 0.32 dB but its optical waveguide length can largely be shortened, and that the M-Z modulator can reach a counteraction of the coherence light when the branching angle is 14.3°. Comparing with common devices, optical splitter and M-Z modulator with low-loss wide-angle Y-Junction have both low loss and small dimension.
In this paper, A new low-loss wide-angle Y-Junction with graded-index optical waveguides, which can change the optical propagation direction, is introduced. An application of the Y-junction structure to 1×4 optical splitter and Mach-Zehnder modulator is analyzed. The two kinds of device performance are examined numerically by the finite difference beam propagation method (FDBPM). The simulated results show that the 1×4 optical splitter has very low loss of 0.32 dB but its optical waveguide length can largely be shortened, and that the M-Z modulator can reach a counteraction of the coherence light when the branching angle is 14.3°. Comparing with common devices, optical splitter and M-Z modulator with low-loss wide-angle Y-Junction have both low loss and small dimension.
2004, 33(6): 643-646,685.
Abstract:
ASON is the next step in the progress of optical trandmission technology. It provides the ability of connection's dynamic setup and release by introducing the intelligence contol plane. The multigranular optical cross connect is the key equipment in ASON.It provides the connection function for optical channel.The two structure of MG-OXC are introduced and compared in this paper.The advantage of MG-OXC comparing to the ordinary OXC is introduced. It's ability to support the control plane of ASON is also introduced.The evultion and trend of waveband and routing algorithm research are introduced in the last part.
ASON is the next step in the progress of optical trandmission technology. It provides the ability of connection's dynamic setup and release by introducing the intelligence contol plane. The multigranular optical cross connect is the key equipment in ASON.It provides the connection function for optical channel.The two structure of MG-OXC are introduced and compared in this paper.The advantage of MG-OXC comparing to the ordinary OXC is introduced. It's ability to support the control plane of ASON is also introduced.The evultion and trend of waveband and routing algorithm research are introduced in the last part.
2004, 33(6): 647-651.
Abstract:
The optical burst switching technology has the advantage of higher efficiency of channel usage and more suitable for packet transfer comparing with the optical wavelength switching. At the edge node, how to arrange the channel resource for burst packets, is the key item for optical burst switching network. In this paper, a wavelengths arrange scheme in optical burst switching edge node is presented, the function of each modules in the scheme and the information flow are introduced also.
The optical burst switching technology has the advantage of higher efficiency of channel usage and more suitable for packet transfer comparing with the optical wavelength switching. At the edge node, how to arrange the channel resource for burst packets, is the key item for optical burst switching network. In this paper, a wavelengths arrange scheme in optical burst switching edge node is presented, the function of each modules in the scheme and the information flow are introduced also.
2004, 33(6): 652-655.
Abstract:
The basic architecture of optical burst switching network core switch node is briefly presented in this paper. The function of switch control module and the structure of core switch node are introduced in brief. Constructing the simulation platform of OBS system is presented. This paper introduces the function of main modules in OBS system, the key technologies involved and the parameters which can be configured. Several structures of OBS core switch node simulation module are built in this paper. Simulation results and analysis of the network performance are also presented. This paper analysis the impaction of burst lost ratio by using different switching structures.
The basic architecture of optical burst switching network core switch node is briefly presented in this paper. The function of switch control module and the structure of core switch node are introduced in brief. Constructing the simulation platform of OBS system is presented. This paper introduces the function of main modules in OBS system, the key technologies involved and the parameters which can be configured. Several structures of OBS core switch node simulation module are built in this paper. Simulation results and analysis of the network performance are also presented. This paper analysis the impaction of burst lost ratio by using different switching structures.
2004, 33(6): 656-658.
Abstract:
In this paper a new buffer management scheme called DT+SMA is presented. This scheme combines sharing with minimum allocation (SMA) and dynamic threshold (DT) arithmetic, and we find it has the benefits of both DT and SMA. Through analysis the scheme in theory and simulation we can see that actually DT+SMA is better than SMA in fairness in single priority, and is easier to apply in multiple priorities model than DT when achieving the similar performance with DT.
In this paper a new buffer management scheme called DT+SMA is presented. This scheme combines sharing with minimum allocation (SMA) and dynamic threshold (DT) arithmetic, and we find it has the benefits of both DT and SMA. Through analysis the scheme in theory and simulation we can see that actually DT+SMA is better than SMA in fairness in single priority, and is easier to apply in multiple priorities model than DT when achieving the similar performance with DT.
2004, 33(6): 659-662.
Abstract:
A novel optical packet generation scheme is proposed based on three branch modulator.Theory analysis and simulation demonstrate that the scheme realizes subcarrier multiplexing label of single sideband and overcomes the nonlinear modulation of Mach-Zehnder modulator on subcarrier signal. The problem of intensity modulation on subcarrier multiplexing label from payload is analyzed.
A novel optical packet generation scheme is proposed based on three branch modulator.Theory analysis and simulation demonstrate that the scheme realizes subcarrier multiplexing label of single sideband and overcomes the nonlinear modulation of Mach-Zehnder modulator on subcarrier signal. The problem of intensity modulation on subcarrier multiplexing label from payload is analyzed.
2004, 33(6): 663-666.
Abstract:
The development of router architecture is analyzed, and the fast route lookup and packet classification algorithms for high speed router are researched. The research shows the lookup algorithm for IPv6 route lookup with compression trie has reasonable memory space and fast lookup time. The compositive multi-dimensional packet classification algorithm based on tree divided by value is scalable. It can deal with prefixes match and range match for large rule sets. Two algorithms are suitable for high speed router.
The development of router architecture is analyzed, and the fast route lookup and packet classification algorithms for high speed router are researched. The research shows the lookup algorithm for IPv6 route lookup with compression trie has reasonable memory space and fast lookup time. The compositive multi-dimensional packet classification algorithm based on tree divided by value is scalable. It can deal with prefixes match and range match for large rule sets. Two algorithms are suitable for high speed router.
2004, 33(6): 667-670.
Abstract:
Optical burst switching network is a long term potential broadband network with new optical switching technology. It is a very important task to study management scheme in OBS network which is an important means to study the characteristics of the network. This paper studies the model of SNMP network management in OBS network, according to which the design of embedded agent is analyzed. Through user-defined serial interface, the embedded agent can monitor local equipments, and through SNMP, it exchanges information with the network management system. And procedure of embedded agent software is provided on UCDSNMP software development package.
Optical burst switching network is a long term potential broadband network with new optical switching technology. It is a very important task to study management scheme in OBS network which is an important means to study the characteristics of the network. This paper studies the model of SNMP network management in OBS network, according to which the design of embedded agent is analyzed. Through user-defined serial interface, the embedded agent can monitor local equipments, and through SNMP, it exchanges information with the network management system. And procedure of embedded agent software is provided on UCDSNMP software development package.
2004, 33(6): 671-673,677.
Abstract:
A scheme for all-optical clock recovery from non-return-to-zero(NRZ) coded data by combining a novel clock enhancement component with a two section distributed feed back(DFB) laser has been proposed, and its' function and performances under 64 Gb/s NRZ data streams have been demonstrated and investigated by numerical simulation. It is shown that this system can achieve clock recovery from high-speed NRZ data with a satisfying performance.
A scheme for all-optical clock recovery from non-return-to-zero(NRZ) coded data by combining a novel clock enhancement component with a two section distributed feed back(DFB) laser has been proposed, and its' function and performances under 64 Gb/s NRZ data streams have been demonstrated and investigated by numerical simulation. It is shown that this system can achieve clock recovery from high-speed NRZ data with a satisfying performance.
2004, 33(6): 674-677.
Abstract:
The self-phase modulation induced optical power transfer function is analyzed in cascaded optical fiber communication system in terms of different chromatic dispersion, arbitrary fiber numbers and lengths. Computer simulations are also done. The results of simulations show that the power change of self-phase modulation (SPM) relates to fiber loss, nonlinear coefficient and modulation frequency of system and fluctuates with change of modulation frequency. The simulations are well accordance with the analytical result.
The self-phase modulation induced optical power transfer function is analyzed in cascaded optical fiber communication system in terms of different chromatic dispersion, arbitrary fiber numbers and lengths. Computer simulations are also done. The results of simulations show that the power change of self-phase modulation (SPM) relates to fiber loss, nonlinear coefficient and modulation frequency of system and fluctuates with change of modulation frequency. The simulations are well accordance with the analytical result.
2004, 33(6): 678-681.
Abstract:
The conception of packet classification and request for it is introduced. The paper classifies the existed packet classification algorithms into three kinds:conversion into single-field search、dependent field search、independent field search, then compare their performance. The presentation of parallel packet classification indicates how to reduce the storage space and update time is the research importance.
The conception of packet classification and request for it is introduced. The paper classifies the existed packet classification algorithms into three kinds:conversion into single-field search、dependent field search、independent field search, then compare their performance. The presentation of parallel packet classification indicates how to reduce the storage space and update time is the research importance.
2004, 33(6): 682-685.
Abstract:
Through numerical simulation the effect of the different dispersion compensation schemes in WDM systems based on FRA are studied depending on the PTDS software. The results indicate that, employing a lumped, under-compensation and post-position compensation scheme is the most optimal.
Through numerical simulation the effect of the different dispersion compensation schemes in WDM systems based on FRA are studied depending on the PTDS software. The results indicate that, employing a lumped, under-compensation and post-position compensation scheme is the most optimal.
2004, 33(6): 686-689.
Abstract:
An improved method on the structure of double sideband phase detection in optical fiber communication system with non return to zero data is described for real-time accumulated dispersion monitoring, the theory of this method is deduced and the numerical simulation is presented. The results show that this method achieves quite precise monitoring value without any modification of transmitters. In the 1.55 μm region, the range of accumulated dispersion for 10Gb/s data reaches ±640 ps/nm.
An improved method on the structure of double sideband phase detection in optical fiber communication system with non return to zero data is described for real-time accumulated dispersion monitoring, the theory of this method is deduced and the numerical simulation is presented. The results show that this method achieves quite precise monitoring value without any modification of transmitters. In the 1.55 μm region, the range of accumulated dispersion for 10Gb/s data reaches ±640 ps/nm.
2004, 33(6): 690-693,705.
Abstract:
In this paper, a hardware implementation scheme of the scheduling module in Optic burst switching(OBS) edge node receiver is presented. Input serial polling(ISP), which is based on virtual output queuing(VOQ) mechanism, is a kind of fair and high-performance algorithm for crossbar arbitrating. Focused on ISP, the design allows receiving, switching and Ethernet-assembling for 6 routs of 1 000M-OBS data with two high-speed FPGA chips. The 6 routs of data mentioned above are totally independent, and there exists communication between two FPGA chips.
In this paper, a hardware implementation scheme of the scheduling module in Optic burst switching(OBS) edge node receiver is presented. Input serial polling(ISP), which is based on virtual output queuing(VOQ) mechanism, is a kind of fair and high-performance algorithm for crossbar arbitrating. Focused on ISP, the design allows receiving, switching and Ethernet-assembling for 6 routs of 1 000M-OBS data with two high-speed FPGA chips. The 6 routs of data mentioned above are totally independent, and there exists communication between two FPGA chips.
2004, 33(6): 694-696,705.
Abstract:
This paper Burst mode synchronization is a key techno logy in passive optical network that utilizes time division multi access (TDMA) with single mode fiber. For the clock recovery in burst mode transmission, a burst mode synchronizer using keyword detection based on multiphase clock is usually used. In this article, based on research in asynchronous transfer moder-passive optical network, author described a method to design burst synchronizer in AS1773 protocol. This synchronizer is implemented by field programmable gate array, which can achieve high performance with low cost, short developing period.
This paper Burst mode synchronization is a key techno logy in passive optical network that utilizes time division multi access (TDMA) with single mode fiber. For the clock recovery in burst mode transmission, a burst mode synchronizer using keyword detection based on multiphase clock is usually used. In this article, based on research in asynchronous transfer moder-passive optical network, author described a method to design burst synchronizer in AS1773 protocol. This synchronizer is implemented by field programmable gate array, which can achieve high performance with low cost, short developing period.
2004, 33(6): 697-700.
Abstract:
How to assemble TCP/IP packets into data burst in OBS edge nodes is discussed. Min-burst-length-max-assembly-period algorithm is preferred and implemented on Xilinx® VirtexⅡ Pro FPGA. A state machine controls the assembly process. Sharing buffer memory is proposed to save memory resources. That is, the number of physical burst queues is less than that of burst types since the inputs is random.
How to assemble TCP/IP packets into data burst in OBS edge nodes is discussed. Min-burst-length-max-assembly-period algorithm is preferred and implemented on Xilinx® VirtexⅡ Pro FPGA. A state machine controls the assembly process. Sharing buffer memory is proposed to save memory resources. That is, the number of physical burst queues is less than that of burst types since the inputs is random.
2004, 33(6): 701-705.
Abstract:
The OBS based on the GMPLS is the trend of current network. The Burst Assembly is the key technology of the OBS based on the GMPLS. It decides whether the whole network can be realized. This paper tells us how to realize the burst assembly by LSP assembly and some GMPLS protocols. In the end, give us that the OBS based on the GMPLS not only resolves the limitations of OBS, but also gives us some new merits.
The OBS based on the GMPLS is the trend of current network. The Burst Assembly is the key technology of the OBS based on the GMPLS. It decides whether the whole network can be realized. This paper tells us how to realize the burst assembly by LSP assembly and some GMPLS protocols. In the end, give us that the OBS based on the GMPLS not only resolves the limitations of OBS, but also gives us some new merits.
2004, 33(6): 706-709.
Abstract:
In this paper, a novel DC coupled feedforward type optical burst mode receiver is presented.Through the cooperation between the fast peak detector and the delay unit,the receiver is realized with lower BER, bettter receiver sensitivity and receiver dynamic range. The output NOR gate can cancell the unwanted peak in dead spaces between packet signals.The results of the simulation indicate that the receiver sensitivity is -24 dBm (average power, 10-9BER) at bit rates up to 1.25 Gb/s and the maximum receivable optical average power is -4 dBm, so the dynamic range is 20 dB. The minimum packet spacing required to properly recover the first bit in the second packet is 20 ns.
In this paper, a novel DC coupled feedforward type optical burst mode receiver is presented.Through the cooperation between the fast peak detector and the delay unit,the receiver is realized with lower BER, bettter receiver sensitivity and receiver dynamic range. The output NOR gate can cancell the unwanted peak in dead spaces between packet signals.The results of the simulation indicate that the receiver sensitivity is -24 dBm (average power, 10-9BER) at bit rates up to 1.25 Gb/s and the maximum receivable optical average power is -4 dBm, so the dynamic range is 20 dB. The minimum packet spacing required to properly recover the first bit in the second packet is 20 ns.
2004, 33(6): 710-713.
Abstract:
According to the idea of wavelength reserving,the application of the state table of wavelength is discussed in this paper. In this way,fair,on-chip resource and frequnece is considered in design of optical burst switch edge node.Through comparation,pipeling technology can be applied in this design successsfully.Finally the paper illustrate implementation of pipeling technology in FPGA device.
According to the idea of wavelength reserving,the application of the state table of wavelength is discussed in this paper. In this way,fair,on-chip resource and frequnece is considered in design of optical burst switch edge node.Through comparation,pipeling technology can be applied in this design successsfully.Finally the paper illustrate implementation of pipeling technology in FPGA device.
2004, 33(6): 714-717.
Abstract:
For the situation of many services in the users' request,the paper introduces DBA algorithm in EPON upstream channel in support of QoS. We propose two control frames and two novel DBA algorithms supporting multi-priority services with differentiated QoS:SGMTA algorithm based on ONU and MGSTA algorithm based on OLT. In our conducted simulation experiments, both algorithm show satisfied delay performance and stable buffer size as the network load increases. Thus, we conclude that these two novel DBA algorithms are reasonable and practical in support of multi-priority EPON system.
For the situation of many services in the users' request,the paper introduces DBA algorithm in EPON upstream channel in support of QoS. We propose two control frames and two novel DBA algorithms supporting multi-priority services with differentiated QoS:SGMTA algorithm based on ONU and MGSTA algorithm based on OLT. In our conducted simulation experiments, both algorithm show satisfied delay performance and stable buffer size as the network load increases. Thus, we conclude that these two novel DBA algorithms are reasonable and practical in support of multi-priority EPON system.
2004, 33(6): 718-721.
Abstract:
This paper discusses the queuing scheme and scheduling algorithms for crossbar switch,and analyzes these algorithms in detail.The similarities and the difference of different algorithms in mechanisms of each class are described,and their performances are compared.Furthermore,the future research directions and possible open problems in this area are discussed.
This paper discusses the queuing scheme and scheduling algorithms for crossbar switch,and analyzes these algorithms in detail.The similarities and the difference of different algorithms in mechanisms of each class are described,and their performances are compared.Furthermore,the future research directions and possible open problems in this area are discussed.
2004, 33(6): 722-725,729.
Abstract:
The thesis introduces the research background and some technical parameters of the route lookup algorithm, and describes some basic algorithms and their pros and cons. Then it presents a fast route lookup algorithm based hardware, six-step route lookup algorithm and its hardware implementation, and the result of its simulation shows it furnishes 50×106 routing lookups/s.
The thesis introduces the research background and some technical parameters of the route lookup algorithm, and describes some basic algorithms and their pros and cons. Then it presents a fast route lookup algorithm based hardware, six-step route lookup algorithm and its hardware implementation, and the result of its simulation shows it furnishes 50×106 routing lookups/s.
2004, 33(6): 726-729.
Abstract:
we have researched on an all-optical 2R regenerator based a semiconductor optical amplifier(SOA). This regenerator is using the self-phase modulation-induced spectral red shift caused by SOA. Through a subsequent optical band-pass filter, we can regenerate the signal.We research on the regenerator in a 10 Gbit/s, 20 Gbit/s and 40 Gbit/s channels of time division multiplex system respectively. We have confirmed that this scheme can be used in time division multiplex system suitably.
we have researched on an all-optical 2R regenerator based a semiconductor optical amplifier(SOA). This regenerator is using the self-phase modulation-induced spectral red shift caused by SOA. Through a subsequent optical band-pass filter, we can regenerate the signal.We research on the regenerator in a 10 Gbit/s, 20 Gbit/s and 40 Gbit/s channels of time division multiplex system respectively. We have confirmed that this scheme can be used in time division multiplex system suitably.
2004, 33(6): 730-733.
Abstract:
The paper puts forward a EPON MAC of fixed size frame structure.The frame structure of fixed size can improve the the process velocity of exchanger.Frame cycle is divided into three segments,ONU not only reports the queue size,but also reports the distribution of frame arrival,Which will decrease the delay jitter of high priority traffic.We adopt the principle of first grant of high priority traffic,this can decrease the access delay of high priority traffic. All can satisfy the requests of real-time voice traffic which is sensitive to the delay and delay jitter.
The paper puts forward a EPON MAC of fixed size frame structure.The frame structure of fixed size can improve the the process velocity of exchanger.Frame cycle is divided into three segments,ONU not only reports the queue size,but also reports the distribution of frame arrival,Which will decrease the delay jitter of high priority traffic.We adopt the principle of first grant of high priority traffic,this can decrease the access delay of high priority traffic. All can satisfy the requests of real-time voice traffic which is sensitive to the delay and delay jitter.
2004, 33(6): 734-737.
Abstract:
Crossbar is used to classify data packets in optical burst switching (OBS) edge nodes. This paper introduces input serial polling (ISP) algorithm, which is usually adopted in crossbar. An algorithm, named variable input serial polling (VISP), is proposed to support variable length data packets assembling. The details of the implementation of VISP algorithm in field programmable gate array (FPGA) is presented. Resource occupation and working frequency are compared between ISP and VISP.
Crossbar is used to classify data packets in optical burst switching (OBS) edge nodes. This paper introduces input serial polling (ISP) algorithm, which is usually adopted in crossbar. An algorithm, named variable input serial polling (VISP), is proposed to support variable length data packets assembling. The details of the implementation of VISP algorithm in field programmable gate array (FPGA) is presented. Resource occupation and working frequency are compared between ISP and VISP.
2004, 33(6): 738-742.
Abstract:
Shared-memory is a common structure of Ethernet switches. It can achieve the optimal throughput and delay performance. Furthermore, for a given loss rate, a shared-memory switch requires less buffers than other switches. This paper analyzes and summarizes the physical and data structure of shared-memory Ethernet switches according to existence shared-memory switch chips, and bring forward our new structure.
Shared-memory is a common structure of Ethernet switches. It can achieve the optimal throughput and delay performance. Furthermore, for a given loss rate, a shared-memory switch requires less buffers than other switches. This paper analyzes and summarizes the physical and data structure of shared-memory Ethernet switches according to existence shared-memory switch chips, and bring forward our new structure.
2004, 33(6): 743-745.
Abstract:
The upper-bounds on BER of OCDMA systems using the time spreading/wavelength hopping codes is investigated with or without using the optical hard limiter.The thorough investigation and analysis on the performances of the systems is performed by simulation and it gets some valuable results. Reducing hit probability between codes and using optical hard limiter can effectively improve the performances of the system.
The upper-bounds on BER of OCDMA systems using the time spreading/wavelength hopping codes is investigated with or without using the optical hard limiter.The thorough investigation and analysis on the performances of the systems is performed by simulation and it gets some valuable results. Reducing hit probability between codes and using optical hard limiter can effectively improve the performances of the system.
2004, 33(6): 746-750.
Abstract:
This paper gives a discussion about different survivability strategies, especially focus on dedicated path-protection(DPP) and shared path-protection(SPP); at the same time, results of simulation experimentation aimed at comparing their performance are analyzed, which fetchs out the conclusion:shared path-protection can use the wave length resources more sufficiently than pedicated path-protection can.
This paper gives a discussion about different survivability strategies, especially focus on dedicated path-protection(DPP) and shared path-protection(SPP); at the same time, results of simulation experimentation aimed at comparing their performance are analyzed, which fetchs out the conclusion:shared path-protection can use the wave length resources more sufficiently than pedicated path-protection can.
2004, 33(6): 751-754.
Abstract:
In this paper, the principle and the architecture of optical burst switching network are introduced. Then the management of the network which is based on simple network management protocol is discussed, including the model and the architecture of the management, the choices of developing tools to manager and agent, the design of management information bases of the core node and the edge ones, and the development of the performance management, configuration management and failure management to the network.
In this paper, the principle and the architecture of optical burst switching network are introduced. Then the management of the network which is based on simple network management protocol is discussed, including the model and the architecture of the management, the choices of developing tools to manager and agent, the design of management information bases of the core node and the edge ones, and the development of the performance management, configuration management and failure management to the network.
2004, 33(6): 755-758.
Abstract:
A systematic classification method aimed at packet-dropping-based buffer management algorithms is present, some dominant algorithms and their variations in today's network are introduced, including some new buffer management algorithms which based on control theories and fuzzy logic, and gives advice on future research for solving the present problems. Because of unique issues in the wireless channel, some modifications are needed when apply these algorithms to wireless network. We present some advices and basic thoughts on this problem.
A systematic classification method aimed at packet-dropping-based buffer management algorithms is present, some dominant algorithms and their variations in today's network are introduced, including some new buffer management algorithms which based on control theories and fuzzy logic, and gives advice on future research for solving the present problems. Because of unique issues in the wireless channel, some modifications are needed when apply these algorithms to wireless network. We present some advices and basic thoughts on this problem.
2004, 33(6): 759-762.
Abstract:
In this paper, we propose a dual threshold admission control algorithm in voice/data integrated code division multiple access cellular networks. The proposed algorithm gives different threshold to data call, new voice call and handoff call. When a call requests, the proposed algorithm estimates the remained bandwidth and compares it with the threshold to decide if the requesting call is accepted. The simulation shows that the algorithm has the different performance with the different threshold values. We choose the adaptive threshold values through the simulation analysis.
In this paper, we propose a dual threshold admission control algorithm in voice/data integrated code division multiple access cellular networks. The proposed algorithm gives different threshold to data call, new voice call and handoff call. When a call requests, the proposed algorithm estimates the remained bandwidth and compares it with the threshold to decide if the requesting call is accepted. The simulation shows that the algorithm has the different performance with the different threshold values. We choose the adaptive threshold values through the simulation analysis.
