The performance of a statistical multiplexer whose inputs consist of a superposition of voice packets from c trunks is studied. Voice calls are established according to a Poisson process and call holding times have a general distribution. Each voice call, while in progress, generates packets at fixed intervals of Δ. Thus, when n calls are in progress, the packet generation process is a superposition of n renewal processes, each with an interarrival time of exactly Δ. These packets are fed to a transmitter whose service time for all packets is exactly β. It is assumed that c β<Δ, i.e. the transmitter queue is stable even when all trunks are busy. The delay for such a system is analyzed by solving the ΣDi/D/1 queue. An efficient hybrid simulation method for the packet voice multiplexer is proposed
A practical method which gives the approximate solution of product form closed queueing networks is developed. The starting point is the integral representation of the normalization constant G. This approach was first used by Mitra et al. (1982). A conceptually simpler solution is proposed. It follows that the memory space requirements are dramatically reduced and the processing time is quite small, even on eight-bit personal computers. An important feature of this approach is that the complexity does not grow with the number of chains (classes), but only with the number of noninfinite server stations.< >
FODA (FIFO ordered demand assignment) a TDMA (time division multiple access) satellite access scheme suitable for simultaneous transmissions of real-time and bulk data is described. It is able to efficaciously counter deep fades due to rain attenuation at high frequencies, always maintaining a good channel optimization. The FODA system, using adaptive bit and coding rates, is designed to obtain the maximum out of a satellite channel in any condition. The system, already operating at 2 Mbit/s with 1/2 coded or uncoded data, is in the middle of partial redesign to cope with fades of the signal due to bad atmospheric conditions when operating in the Ka band. The reaction of the system to the fade effect is shown. The results of a deterministic study on frame utilization when fades must be countered are also presented.< >
The performance of convolutional codes combined with continuous phase modulation schemes (CC/CPM) with M-algorithm decoding is evaluated by computer simulation. New CC/CPM schemes with a feedback convolutional encoder having, in some cases, better distances than previous CC/CPM which uses a feedforward encoder, are presented. Comparison of CC/CPM and CPM, constrained to the same complexity and bandwidth, shows that the former have a better performance. The simulation results also corroborate the fact that the M-algorithm is far more efficient than the Viterbi algorithm
A packet switch based on electrooptical technologies is presented. It is composed of N VLSI chips which interface the input and output links with the internal optical interconnection network. Its simulation, when loaded with homogeneous input and output traffic, showed that two packets at a time must be first extracted from each input buffer and then transmitted on the optical interconnection network to achieve an almost optimal performance. This significantly affects both the architecture of the VLSI chips and the optical transmission scheme.< >
An algorithm is developed to analyze the dissipation of electromagnetic power in waveguide of arbitrary cross section, based on the finite-element method and the Galerkin principle. This procedure is applicable to inhomogeneous media characterized by their loss tangent and resistivity. The wave equation, which is complex due to the presence of dissipative media, is discretized into a complex matrix eigenvalue problem. This matrix system is solved by an iterative procedure with the aid of the Newton-Raphson method, reducing the complex matrices simultaneously, one to upper Hessenberg form and the other to upper triangular form. To verify the accuracy of this method the fundamental mode of propagation in a rectangular waveguide with conductive walls with a slab of dissipative material placed along the symmetry plane of waveguide is analyzed. An image line formed by a silicon rod with two different values for the loss tangent (5.0×10-4; 1.0×10-3) is also studied
Two multiple-access protocols for mobile data services on error-prone transmission channels are studied. TDMA (time division multiple access) and a class of demand assignment techniques (DAMA) are compared using the same user and channel models. TDMA is analyzed by a straight-forward computation. and for the demand assignment protocols a model consisting of two embedded Markov chains is developed. Channel utilization and packet-delay are evaluated. It is shown that the advantages of DAMA in error-free channels may diminish with increasing error rates
It is pointed out that the interworking between the ISDN (integrated services digital network) user-network interface layer 3 protocol and the SS7-ISUP (ISDN user part) protocol is not straightforward, in spite of the fact that ISUP was developed to support ISDN. Three basic criteria for designing two protocols, such that their interworking is simplified, are presented, and it is shown that the ISDN-ISUP interworking does not follow these criteria
The development of high speed networks with data rates of 100 Mbit/s and more as well as the design of advanced applications imposes new requirements on the performance of network nodes. The author describes an approach to overcome the upcoming processing bottlenecks of communication protocols. The approach is based on the use of multiprocessor architectures. Several parallel concepts inside network nodes such as pipeline and array constructs are outlined, and different memory concepts are suggested. Performance measurements of protocol implementations on transputer networks following the proposed parallel concepts indicate very promising performance results.<>
A channel separation algorithm for the frequency division multiple access/time division multiplexing (FDMA/TDM) scheme is presented. It is shown that implementation using this algorithm can be wore effective than the fast Fourier transform (FFT) algorithm when only a small number of carriers need to be selected from many, such as satellite Earth terminals. The algorithm is based on polyphase filtering followed by application of a generalized Walsh-Hadamard transform (GWHT). Comparison of the transform technique used in this algorithm with discrete Fourier transform (DFT) and FFT is given. Estimates of the computational rates and power requirements to implement this system are also given.
A novel complex-input-complex-output transmultiplexer (TMUX) algorithm is derived with application to quadrature modulation schemes for use in the area of satellite onboard processing. The proposed approach exploits an inherent property of the all-zero lattice filters whereby two reverse polynomial all-zero filters are simultaneously derived. This property is coupled to a similar property requirement of the linear phase prototype filter-based doubly odd transform-polyphase filter TMUX, and the resulting computational load is analyzed for the European study concerning multiple carrier demodulation (MCD) equipment. The lattice subfilter algorithm offers a more robust solution to the transmultiplexer implementation than does the transversal scheme when fixed point digital signal processing is applied
The database developed for the meteorological radar of Spino d'Adda in north Italy is presented as an example of automation in this field. The system has facilities for collection, retrieval, and modeling of radar images of rain as well as their components (individual patterns) and descriptors. The SPINO80-GRM software applies techniques of image analysis to radar rain data in order to identify and characterize the required data items (rain patterns) and then organizes them into a hierarchical, relational database. The component retrieval is based on a free combination of its attributes: some statistical investigations are also possible.< >
The adaptive multilevel code (AMC) is a DC-free line code with four levels and two lengths. Both levels and lengths carry information. For a stationary memoryless source, closed-form expressions for the spectra of the AMC coded signals have been obtained analytically using the transition probability matrix method. Results comparing the one-sided spectral densities of the AMC, alternate mark inversion, and biphase waveforms are presented. The advantages of AMC are demonstrated
Modulation via Fourier transform division multiplexing (FTDM) has been suggested for transmission over channels whose signal-to-noise ratios, as functions of frequency, are far from flat. Original designs following the classical water-pouring theorem left a large gap between the theoretical promise of the method and the actual rates achieved. The author presents heuristic input-power allocations for enhancing the performance of such schemes. He also considers FTDM under peak-power constraints
The authors present an architecture and implementation for an FDDI (fiber distributed data interface) concentrator with media access control (MAC) sublayer support. The architecture calls for a distributed implementation of the SMT CMT (station management connection management) for the concentrator. The main features include microcontroller-based physical connection management, configuration management, and link error monitor for the back-end and front-end PHYs. Other features include a node-processor-based entity coordination management, ring management, MAC support, SMT frame transaction support, and link layer control. Communication, between the node processor and the peripheral microcontrollers is two-way serial for high-speed support of concentrator management transactions between the physical entities and the node processor. The architecture is modular because the microcontroller-based management allows for local intelligence for a group of back-end physical entities
A land mobile satellite system (LMSS) experiment was performed within the framework of ESA's PROSAT phase II program in order to analyze the dominant propagation factors in Europe. The measurements were made using a van as the mobile platform and the MARECS satellite as the 1.5-GHz signal source. Various environments in Spain, France, and Sweden were selected. Similar tests took place along the Spanish Railways after mounting the equipment in a locomotive. It is shown that, although shadowing is the main signal impairment to be considered, it does not prevent communication in most cases. The satellite signal can be received in relatively good conditions with a rather high probability in almost any environment in the selected areas of Europe. Even in the worst cases (Stockholm, dense forest), it is possible to find stretches of road where the signal level is high
A simulation model of multipath propagation is used to characterize the relationship between the cross-polarization discrimination and the co-polar attenuation in frequency-reuse line-of-sight links. The behavior of several links are then simulated, and the results are compared with measurements, to test the prediction capability of the model
A method is presented in which a computer is used to randomly simulate the statistical distributions of frequency selective fading. Relationships between the in-band frequency correlation coefficient and antenna height, the ground reflection coefficient, and the frequency bandwidth are randomly simulated. Using this method, frequency selective fading was randomly simulated over the Pacem 2 experimental path (a distance of 100 km, operating in the 11.123 GHz to 11.187 GHz range) from Bouffry to Viabon for the total fading period of 1985
The authors present an original and simple method of speaker-independent isolated work-recognition applied in the recognition of numbers (from 0 to 9) spoken in Portuguese. The recognition process depends on robust parameters and the identification of vowels inside the word under test, both of which are speaker-independent. The main advantage of this method is the fast response time, when compared with the usual methods of recognition, such as template matching algorithms. This method is implemented in a C language algorithm. The recognition rate achieved is about 90%
An algorithm for calculating lower bounds to the performance of all size-constrained vector quantizers has been derived earlier. Applications to the AR(1) source have shown that these lower bounds are very close to the actual rate-distortion curves, even for small sizes of the output alphabet. In this algorithm, however, the codebook changes with rate. In a number of applications it is desirable to operate with the same codebook. The authors modify the above algorithm to calculate lower bounds to the performance of all such multirate coders. An important conclusion is that constraining the codebook vectors does not degrade performance as compared to the size-only constrained case. The authors have proposed a practical alphabet and entropy constrained vector quantizer (AECVQ). This coder is obtained by modifying the ECVQ algorithm so that the codebook remains the same for all rates.< >
