For a spectrum-efficient, bandwidth-on-demand multiplexer-demultiplexer for a multiuser, multicarrier communication system based on wavelet packet trees for downstream transmission, an innovative scheme for synchronization is proposed which uses a unique sync word at the root of the tree. An example filterbank tree with 8 leaves is used throughout the paper. For this example the creation of the unique 32-bit sync word is described. The multirate signal processing and construction and properties of 25 32/spl times/32 matrices required in the multiplexer to find locations of sync words at input ports to cancel the effects of data and insert the 32-bit sync word in a window in the correct location at the root of the tree is described. The ratio of maximum to minimum absolute values of eigenvalues of the matrices is used to ensure well-conditioning, since they must be inverted in the processing.

A set of desirable characteristics of a multicarrier spectrum-efficient bandwidth-on-demand multiplexer-demultiplexer pair for use in mobile satellite and personal communication systems is identified and described. New characteristics are the use of single VSB channels, design of multiplexer channels based on wavelet packet trees which have specified stopband attenuation, overlap of the multiplexer channel magnitude frequency responses at the 3-dB points for spectral efficiency, bandwidth on demand, reasonable lengths for the overall equivalent filters for each multiplexer-demultiplexer channel from input to receiver output, and low-complexity receivers. A multirate digital transmultiplexer is proposed consisting of a wavelet packet-based synthesis filter bank tree followed by a DFT polyphase synthesis filter bank at the transmitter, and a matching demultiplexer at the receiver. A simplified receiver for reception of one channel at a time is described. BER performances when there are phase and timing errors are given.

The relationship of filter bank trees to wavelet packet trees for subspaces, decimation operators, and coordinates of the projections of a function onto the subspaces is briefly explained. Two designs for the filter banks are summarized and the equations for the designs are provided. For one of these the prototype filters are all the same, and for the other they are the same at each level of the tree. The best method for spectral factorization of high-order polynomials, needed for the filter designs, is selected. Lattice structures preserving perfect reconstruction with finite word lengths are used in computer simulations to study quantization effects.

Systematic design methods for multifilters and multiwavelets are presented. Explicit formulae are given for aliasing cancellation in biorthogonal multifilter banks. The relationship with vector transforms over the real or finite fields is described.

Design methods for regular multidimensional perfect reconstruction (PR) filter banks are described. A systematic method for two-channel and four-channel filter banks is presented. The main novelties are: (1) the filters have impulse response with square support, rather than diamond support, (2) regular designs that have rectangular support are also presented, which are highly efficient in practice, since expensive memory is saved; (3) in addition to new diamond filters for the two-channel case, hexagonally-symmetric filters are also derived; and (4) novel 3-D filter banks are also designed. In all cases the filters are linear phase, achieve arbitrarily high regularity and can be used to obtain biorthogonal wavelet bases. The filter banks can be implemented in a structurally perfect-reconstruction manner.

A communication-theoretical model adapted to ATV (advanced television) and HDTV (high definition television) systems, partitioned into functional blocks to which performance criteria and theoretical limits can be applied, is briefly described. The types of signal processing which are or could be employed in each block to decrease the effects of channel degradations are described for both analog and digital transmission, with primary emphasis on analog transmission for terrestrial broadcast. Digital transmission over these channels is treated briefly. Analog and digital transmission over satellite and fiber optic channels is summarized. New signal processing functions are discussed.>

The explicit form of the uncertainty principle for the 3-D signals of advanced television (ATV) is given, and the conditions for efficient use of spatiotemporal BW are discussed. Applications to extrapolation and to approximation in an interval by a bandlimited 3-D function are summarized. A theory for spatiotemporal interchange based on the analog between the inertia tensor of classical mechanics and a dispersion tensor for a 3-D image is described and a possible approach to spatio-temporal interchange is presented. Potential applications of previous results in image processing to location and resolution of objects and coding are discussed.>

The technical advantages of measuring complex impulse responses rather than the in-phase component of broadcast teletext channels and the extraction of channel information from real impulse response measurements and its limitations are described. A number of illustrative examples are provided. It is concluded, after discussion, that measurements of channel complex impulse responses are necessary, in spite of extra cost, so that suitably flexible and general computer simulation design and performance studies of broadcast teletext systems can be made.>

To accurately predict the performance of broadcast teletext systems, one must know the characteristics of typical over-the-air broadcast television channels. Consequently, ensembles of typical complex impulse responses of broadcast teletext channels in the VHF and UHF frequency ranges are required for computer performance simulations. To circumvent the lack of comprehensive measured channel data, a theoretical multipath propagation model for suburban and urban environments was developed to gain insight into the nature of typical television multipath channels. In this model, both the direct and scatter paths consist of a line-of-sight and ground-reflected component. The basis of the propagation model is an expression for the ratio of the scattered path to direct path field strength caused by vertically oriented dielectric slabs of specified thickness, height, and width. The reflector location, orientation, width, height, and thickness are randomized for each Monte Carlo simulation cycle. A scatter table of relative amplitude versus multipath delay is maintained for subsequent analysis and formation of typical multipath channel impulse responses. A perceived desired-to-undesired ratio (PDUR), which is essentially a signal-to-clutter ratio, is computed to eliminate channels of unacceptable video quality, as these would be of no interest to the teletext application. A modification to account for near-field scattering is described. Finally, mobile channel applications are discussed.

Error correction can greatly improve the performance and extend the range of broadcast teletext systems. In this paper, the requirements for an error-correcting scheme for broadcast teletext in North America (NABTS) are set down. An error-correction scheme which meets all these requirements is then described. The simplest case employs the one parity bit in each 8 bit byte and no suffix of parity check bits at the end of each data block. The next level also uses a single byte of parity check bits at the end of each data block. Adding a second byte of parity checks at the end of each data block results in a Reed-Solomon code, called theCcode, for each data block. Adding one data block of parity checks afterh - 1data blocks results in a set ofhdata packets being encoded into a bundle, in which verticalCcodes provide powerful interleaving. In a final alternative, two data blocks hold the check bytes for the vertical codewords, and the most powerful coding scheme, the double bundle code, results. The detailed mathematical definitions of the various codes are referred to or described, formulas for performance calculations are referred to, and performance curves are presented for the AWGN channel as well as for measured field data. These performance curves are discussed and compared to the performance of a difference set cyclic code, originally designed for the Japanese teletext system, which corrects any 8 bits in error in a packet.

An isomorphism is derived between a typical bandpass broadcast television system to be used in transmitting teletext data and its complex baseband equivalent representation for use in analysis and computer simulations. The distortions introduced by a multipath channel are discussed. The complex noise process is modeled by a mixture of Gaussian and impulsive noise, and the parameters of the noise model are derived. The video signal-to-noise ratio SNRTVis related to other parameters such as the input noise variance, output noise power, and the carrier-to-noise density ratio. A complete simulation model for broadcast television transmission and reception is shown in block diagram form. The remaining components of the computer simulation model are briefly discussed and references given. These include topics such as amplitude control, computer generation of ensembles of complex impulse responses, simulation of synchronous, quasi-synchronous and envelope detectors, bit slicers, bit synchronizers, error-correcting codes, estimation of error statistics and bit error rates, and adaptive equalization. Validation is also briefly noted with references. Several examples of simulation results are given. Some of the implementation details, simulation results, and a flow diagram of the simulation program are provided. This teletext computer simulation model appears to be the most complete one created to this time and has been used extensively in performance analyses to evaluate new designs and gain insights into design improvements.