JP2009513064A - Apparatus and method for wireless transmission of uncompressed video - Google Patents

Abstract

【Task】
An apparatus and method for wireless transmission of uncompressed HDTV video overcomes the problem of transmitting enormous amounts of information over a wireless link. This is realized by directly mapping the conversion factor of the video element to a communication symbol such as an OFDM symbol. The main part of the important transform coefficients, for example the MSB of the coefficients representing the low frequency of each video element, in particular the quantized values of the low frequency elements, are transmitted in a coarse representation, for example using QPSK or QAM . The coefficients representing the high frequency of the video elements, the quantization error values of the low frequency elements, or their non-linear transformations, are transmitted as a pair of complex real and imaginary parts with points in a fine arrangement. The present invention further provides a delay and bufferless embodiment of the transmitter and receiver pair.
[Selection figure] None

Description

  The present invention relates to the transmission of uncompressed video over a wireless link. In particular, the present invention relates to delay and bufferless transmission of uncompressed HDTV video over a wireless link using direct mapping of image transform coefficients to transform symbols.

  In many homes, television and / or video signals are received by fixed set-top boxes in the home via cable or satellite links. In many cases, it is necessary to place the screen several meters away from the set top box. This trend is spreading as flat screens using plasma or liquid crystal display (LCD) televisions are hung on the walls. Typically, connecting a screen to a set-top box via a cable is not preferred for aesthetic reasons and / or ease of installation. Therefore, wireless transmission of the video signal from the set top box to the screen is preferred. Similarly, it is desirable to locate a computer, game controller, VCR, DVD or other video source that produces the video displayed on the screen at a location remote from the screen.

  Usually, for example, data compressed in a motion picture expert group (MPEG) format is received by a set top box and decompressed into a high quality unprocessed video signal by the set top box. This unprocessed video signal may be in an analog or digital format such as a digital video interface (DVI) format or a high resolution multimedia interface (HDMI) format. These digital formats typically have high-definition television (HDTV) data rates of up to about 1.5 gigabits per second (Gbps).

  Wireless short-range transmission within a home can occur over an unauthorized band of about 2.4 GHz or about 5 GHz (eg, the 5.15-5.85 GHz band in the United States). These bands are currently used in wireless local area networks (WLANs), where the 802.11 WiFi standard is 11 Mbps (802.11b) or 54 Mbps (20 MHz band and the 802.11g / 802.11a standard). ) Allows the maximum data rate. By utilizing a new multi-input multi-output technology, the new 802.11n standard data rate is over 200 Mbps when the 20 MHz band is used, and double that when the 40 MHz band is used. Can be increased. Another alternative is to use ultra wideband (UWB), which requires providing 100 to 400 Mbps.

  Since the available data rate is lower than the 1.5 Gbps required for uncompressed HDTV video, the video typically needs to be recompressed for wireless transmission as needed. For example, known strong video compression means including 1:30 or more compression elements require very complex hardware to perform the compression. This is usually not suitable for home use. These compression means usually use, for example, wavelet, discrete cosine transform (DCT) or Fourier transform to transform the image into different domains and perform compression in that domain. Conversion usually does not associate data and allows efficient compression. In PCT application IL / 2004/000779, Wireless Transmission of High Quality Video is granted to co-applicants and is hereby incorporated by reference in its entirety, which shows how to transmit a video image. The method includes providing a high-resolution video, encoding each video based on neighboring pixels and compressing the video using an image domain compression method, and fading the compressed video to a transmission channel. And transmitting via.

  Obrador's US Patent Publication 2003/002582 describes wireless transmission of images encoded using joint source channel coding (JSCC). The transmitted image is stretched into multiple subbands of different frequencies. A low resolution image and corresponding boundary coefficients are transmitted. An exemplary JSCC applies channel coding techniques to source code coefficients, protecting relatively important low frequency coefficients, and protecting less important non-critical high frequency coefficients. Another technology of JSCC was proposed by Ramstad at The Marriage of Subband Coding and OFDM Transmission, Norwegian University of Science and Technology (July 2003), which includes, for example, image and OFDM (orthogonal frequency division multiplexing) modulators. Integrate source subband coding.

  In the digital transmission means, the signal is transmitted in a symbol format. Each symbol can have one of a predetermined number of possible values. The set of possible values for each symbol is referred to as a constellation, and each possible value is referred to as a constellation point. The distance between adjacent points affects the immunity to noise. Noise causes the reception of another point instead of the intended point, which causes the symbol to be misinterpreted. In an orthogonal frequency division multiplexing (OFDM) communication scheme, a symbol is composed of a plurality of bins in the frequency domain, for example, 64, 128, or 256 bins, and each domain of each symbol is configured in a two-dimensional arrangement. It is well known in the art that the use of some of the available bins is not recommended. There are usually bins located at the end of the transmission band. Typically, for example in 802.11a / g, 16 of 64 available channels are not used, thus reducing band efficiency.

  In the US patent applications 2004/0196920 and 2004/0196404 of Loheit et al., Another scheme for transmission of HDTV over a wireless link is proposed. The described scheme transmits and receives uncompressed HDTV signals over a wireless RF link that provides a clock signal that is synchronized with the uncompressed HDTV signal. The scheme also includes a clocked data recovery module that provides a stream of data regenerated from an uncompressed HDTV signal. The duplexed data stream is separated into an I data stream and a Q data stream. A modulator connected to the duplexer modulates a carrier having an I data stream and a Q data stream. According to Loheit et al., RF links operate in a variety of frequency bands from 18 GHz up to 110 GHz, thus requiring sophisticated and more expensive transmitters and receivers.

  In view of various limitations of the prior art, it would be beneficial to provide a solution that would allow reliable wireless transmission of HDTV while not requiring aggressive or complex compression or complex hardware implementation. is there. In particular, it is beneficial to avoid compression that depends on having a frame buffer to reach the compression level needed to transmit the vast amounts of data required in applications such as wireless transmission of HDTV data streams. . Furthermore, it is beneficial to avoid the use of very high frequencies in order to achieve the purpose of wireless transmission of HDTV data streams. This is also beneficial if the proposed system does not insert a delay in the transmission of the video. Furthermore, this is beneficial if more efficient use of the transmission band is realized, thereby allowing transmission of more information.

  Uncompressed wireless transmission apparatus and methods overcome the problem of transmitting vast amounts of information over wireless links. This is realized by directly mapping the conversion coefficient of the video element (for example, Y-Cr-Cb element) to a communication symbol such as an OFDM symbol. The main part of the important transform coefficients, for example, the most significant bits of the coefficients representing the low frequency of each video element, in particular the quantized values of these DCs and approximate DC elements, use eg QPSK or QAM. Transmitted as a coarse digital display. Coefficients representing the high frequency of each video element, quantization error values of DC and approximate DC elements, or their non-linear transformations, are transmitted as a pair of complex real and imaginary parts containing points in a fine arrangement of precision. . Thus, the present invention provides a bufferless embodiment with no transmitter and receiver pair delays.

  The disclosed invention provides a scheme that enables transmission of video over a wireless link, such as a high definition television (HDTV), using transmission symbols such as symbols of an OFDM scheme, thereby solving the problems of the prior art. Intended to solve the problem. In particular, the present invention can map a block of pixels or some coefficients directly to a transmitted symbol after a correlating transformation. This decorrelation transformation is performed for the purpose of minimizing the energy of the coefficients without compromising the available degrees of freedom. For example, discrete cosine transform (DCT) is performed on each pixel block of Y, Cr and Cb elements of the video. The Y element provides pixel brightness, while the Cr and Cb elements provide color difference information known as chrominance. In the preferred embodiment, all coefficients are transmitted according to the disclosed invention. In another embodiment of the disclosed invention, only some of the coefficients are used for transmission purposes, thereby preventing very high spatial frequency coefficients and maintaining low spatial frequency coefficients. . Many of the coefficients associated with Y are significantly more protected for wireless transmission than the other two factor coefficients because the human eye is more sensitive to luminance than chrominance. For purposes of illustration and not limitation, a ratio of at least three coefficients for each Y element, such as a 3: 1: 1 ratio, is used for each Cr and Cb element. However, other ratios may be used without departing from the spirit of the invention. Therefore, in the disclosed invention, the DC coefficient equal to or higher than the coefficient representing the high frequency and the approximate DC coefficient are emphasized, and each coefficient of luminance equal to or higher than each coefficient of chrominance is appropriately processed. Unlike compression techniques such as JPEG and MPEG, the present invention further transmits quantization error information over a transmission channel, thereby enabling reconfiguration of the video frame, via a transmission channel, wire or wireless. Provide transmission of essentially uncompressed video, particularly high resolution video.

  In the present invention, a DC coefficient or a DC approximation coefficient having a large value is represented by a rough method sometimes referred to as digital, that is, a part of the DC value is regarded as one of arrangement points of a plurality of symbols. It is represented. This is achieved by quantizing these values and mapping the quantized values according to the disclosed invention. In addition to high frequency coefficients, the quantization error of the DC and DC approximation elements whose main part is displayed coarsely is paired, providing a highly accurate, almost analog value that provides a continuous display of these values in extreme detail Each of these pairs is arranged as a real value and an imaginary value of a complex number that provides

ここで、ｘは係数の値であり、αは、ｆ（ｘ）の指数（power）をｘの指数と同じにするように設定される。 Optionally, a non-linear transformation called compression / decompression is performed on any of these values including complex numbers. Compression / decompression is a non-linear transformation of values. Common compression / decompression functions are logarithmic, for example, A-law (A-law) and μ-law (μ-law). Utilizing these techniques effectively provides a more dynamic range and a more signal to noise ratio in the display of corresponding values. In the preferred embodiment, the following compression / decompression function is used.

Here, x is a coefficient value, and α is set so that the power of f (x) is the same as the power of x.

ここで、αは、ｆ（ｘ_１，ｘ_２）の指数をｘ_１およびｘ_２の指数を合わせた指数に維持するように設定された因数である。開示されている発明の一実施例では、変換されたデータは、符号化してもよい。 Another feasible mapping can map the number of data values to a smaller number of values, thereby saving transmission bandwidth. For example, two numbers are mapped to one number, or three numbers are mapped to two numbers and the like. A particular distortion is inserted when the original value is reconstructed, and the advantage is that it can send unimportant data over a limited available bandwidth. Thus, two exemplary numbers, eg, x ₁ and x _2, are transformed by a function that yields a single value (x ₁ , x ₂ ), where (x ₁ , x ₂ ) is As shown, a factor α set to make the index after mapping the same as the index before mapping is further multiplied.

Here, α is a factor set so as to maintain the index of f (x ₁ , x ₂ ) as an index obtained by combining the indices of x ₁ and x ₂ . In one embodiment of the disclosed invention, the transformed data may be encoded.

  The invention disclosed herein, such as MPEG, can maintain all the coefficients of the decorrelation transform. For this reason, the reconfiguration on the receiving side is more accurate because more information is available for such a reconfiguration. In addition, in the disclosed invention, the subchannels of the transmission channel, which are usually avoided to provide the necessary margin or to prevent interference problems, transmit the values of these coefficients to receive a smaller indication. Can be used for the purpose. By transmitting insignificant values determined by the disclosed invention beyond subchannels or subbands that are not normally used, the bandwidth available for transmission is substantially increased. Furthermore, some values can be made compact by using the method described above.

  All coarse and fine placement points generated as described above are configured as a series of complex numbers adjusted for transmission. In the preferred embodiment of wireless communication, orthogonal frequency division multiplexing (OFDM) is used, but not limited to the disclosed invention. In the OFDM communication scheme, a symbol is composed of a plurality of bins in the frequency domain and each bin (complex number) of each symbol configured in a two-dimensional arrangement. A communication system having a bandwidth W has 2 W of freedom. If the spectral coefficient ρ is smaller than 100%, the number of degrees of freedom is 2 Wρ / sec. Since each complex number includes two degrees of freedom, the number of complex numbers that can be transmitted is ρW. By using multiple receive antennas, ie, multiple transmit antennas that require multi-in, multi-input, multi-output (MIMO), the transmission rate for a given band is increased.

  A detailed description of the principles of the disclosed invention follows. Although the present invention has been described with respect to particular embodiments and drawings, it is not intended to limit the scope of the invention, but is provided for illustrative purposes only.

  FIG. 1 is an exemplary, non-limiting block diagram of a system 100 for directly encoding symbols in accordance with the disclosed invention. The system 100 receives a red-green-blue (RGB) element of a video signal, eg, an HDTV video signal. The RGB stream is converted into a luminance element Y and two color difference elements Cr and Cb in the color conversion block 110. This conversion is well known to those skilled in the art. In one embodiment of the disclosed invention, the video begins with a Y-Cr-Cb video signal when the color conversion block 110 is not needed. The Y-Cr-Cb element is input to the transform unit 120, where a decorrelation transform is performed on each pixel block of each of the three elements. In one embodiment of the disclosed invention, block 120 performs DCT on the pixel block. The pixel block may include 64 pixels arranged in an 8 × 8 format as shown in FIG. The conversion unit 120 may comprise a single subunit that performs the desired conversion, eg, DCT, and processes the conversion of all pixel blocks of the entire video frame of each Y-Cr-Cb element. In another embodiment, a dedicated conversion subunit may be used for each Y-Cr-Cb element, thereby improving system performance. In yet another embodiment, multiple subunits are used, and two or more such subunits capable of performing the desired transformation on the pixel block are used for each Y-Cr-Cb element, Thus, the performance of the system 100 may be further improved. The output of the conversion unit 120 is a series of coefficients input to a mapper 130. Mapper 130 selects these coefficients from each Y-Cr-Cb element transmitted over the wireless link. The mapper 130 also maps the coefficients that are sent to the transmitted symbols, eg, symbols of an orthogonal frequency division multiplexing (OFDM) scheme, and this process is described in more detail in connection with FIG. The symbols are transmitted utilizing an improved OFDM transmitter 140 that handles the mixing characteristics of symbols having a mixture of coarse and fine placement values, as will be described in more detail in FIG. In one embodiment of the disclosed invention, the improved OFDM transmitter 140 is connected to multiple antennas in order to support a multi-input, multi-output (MIMO) transmission scheme, thereby substantially Increase bandwidth and transmission reliability. Further, those skilled in the art will recognize a receiver, eg, the receiver shown in FIG. 8 configured to receive a radio signal comprising symbols transmitted in accordance with the disclosed invention, with a coarse and fine display of transmitted symbols. It can be seen that each coefficient is reconstructed and the inverse transformation of the reconstruction of the Y-Cr-Cb element is performed. However, since there is no frame-to-frame compression, this system does not require a frame buffer. Mapping and transformation are fast, do not need to consider extensive image correlation, frame-to-frame correlation, and work with small blocks, so there is virtually no delay for the operations disclosed here, , Only a limited number of lines are kept in the frame processing. Elements and receivers are described in more detail below.

  In the disclosed invention, decorrelation transformations such as DCT are performed on pixel blocks, eg 8 × 8 pixels, Y-Cr-Cb elements of the video. A two-dimensional coefficient matrix 220 is generated as a result of the transformation of the block, eg, block 210 shown in FIG. The coefficients close to the origin in region 222 are typically the low frequency and DC portions of each Y-Cr-Cb element, such as coefficient 222-i. Higher frequency coefficients, such as coefficients 224-i, 224-j, and 224-k, are present in region 224 and are typically much smaller than the DC element, eg, less than half the size of the DC element. Even high frequencies may be in the area marked as 226. The inventor has noted that high frequencies in region 226 can be removed for Y-Cr-Cb elements to maintain an essentially uncompressed video. Region 226 may be smaller or larger depending on the number of coefficients transmitted in a particular implementation. The main part of the DC element, for example, the most significant bit of the coefficient 222-i is preferably mapped to one of a plurality of arrangement points as shown in the arrangement map 230. The placement map may be 4QAM (QPSK), 16QAM, or any other suitable type. Since four placement points 231 through 234 are shown in the placement map 230, the implementation of 4QAM is taught in this example, and each placement point is mapped with a digital value 00 through 11, respectively. The quantized value of the coefficient 222-i is mapped to one such constellation point depending on the specific value. Such a mapping can be considered as a mapping of digital values.

  However, this coarse representation may have a value that depends on the quantization error, ie the difference between the original value and the value represented by the coarse representation. This error basically corresponds to the least significant bit of the quantized significant coefficient value. The quantization error value is mapped as a real part of the complex number constituting the constellation point 240-i, eg, symbol 240-i, as well as a coefficient that is not represented in a coarse manner, ie, a coefficient associated with a high frequency in region 224. May be. For example, the coefficients 224-i and 224-j may be mapped to the imaginary part and the real part of the placement point 240-j. This allows a substantially continuous display using available points of the placement map or by providing fine placement. Such a mapping is considered a continuous value mapping.

As described above, a receiver capable of receiving the symbol stream disclosed herein, such as the receiver shown in FIG. 8, should be able to reconstruct the coefficients from the transmitted symbols, as described in more detail below. Is done. The invention of MIMO with continuous display and OFDM with continuous display offers advantages over the prior art. In particular, only a simple and simple algebra calculation is necessary to receive the fine values of the transmitted video stream. Even if an error occurs, the impact on the quality of the video stream is very limited and usually not observable. On the other hand, MIMO and / or OFDM systems that transmit pure data, including video transmitted as data rather than by the method disclosed in the present invention, require greater processing power, greater frequency bandwidth, and are usually used immediately. It is not possible and the image quality is sensitive to errors.

  An exemplary reference is shown in FIG. 3, where an 8 × 8 coefficient matrix is assumed, so there are 64 coefficients for each Y—Cr—Cb element. However, for the reasons discussed above, typically 28-64 Y element coefficients, 12-64 each Cr and Cb element coefficients are transmitted over the wireless link. The exact number of coefficients is determined based on the available number of OFDM symbols with each bit of the OFDM symbol having two degrees of freedom available for wireless transmission, and based on the desired level of wireless transmission reliability May be. A 20 MHz OFDM channel allows for up to 20M complex numbers, 20M real numbers per second and 20M imaginary numbers, ie 40M degrees of freedom per second. In MIMO, which substantially expands the available bandwidth, a system with four transmit antennas and four such 20 MHz OFDM channels is available, so theoretically every second. A maximum number of 160M or degrees of freedom is possible. In practice, full spectral efficiency cannot be achieved. With the disclosed technique, the spectral efficiency of the disclosed solution is typically ~ 75%, so in the example described, each transmission channel has about 30M degrees of freedom per second, or 1 Transmit every 120 seconds with a total freedom of 120M. In the disclosed invention, some channels are used to transmit a coarse display and the remaining channels are used to transmit a fine display. More important coefficients are provided with more degrees of freedom, while less important coefficients or their quantization errors are provided with less degrees of freedom. In an exemplary transmission of HDTV video, a single frame includes approximately 1200 OFDM symbols corresponding to 256 bins and includes approximately 14,400 blocks of 8 × 8 pixels. By using a 40 MHz bandwidth channel, twice as many coefficients can be transmitted, more coefficients can be transmitted more accurately, eg, by repeating information in a coarse transmission, coarse information that is more important in a coarser way Can be sent.

  FIG. 4 shows a schematic flow chart illustrating the principle that is central to the disclosed invention. In S410, the video stream is subjected to decorrelation conversion. As a result, a plurality of coefficients describing the components of the original video stream are provided. In S420, the DC and DC approximation coefficients are selected for a coarse display for transmission according to the disclosed invention. In the disclosed invention, the coarse display is quantized in S430, and the coarse display quantized in S440 is mapped to the symbol. In one embodiment of the disclosed invention, the high frequency coefficient portion is used in conjunction with a fine display stream. In yet another embodiment of the disclosed invention, as will be described in more detail below, fine display data or portions thereof are nonlinearly transformed. In S460, the fine display values are mapped to pairs of real and imaginary parts of the symbol. Finally, in S470, the generated symbols are transmitted according to the disclosed invention.

  FIG. 5 shows an exemplary and non-limiting flowchart 500 of processing HDTV video for wireless transmission utilizing an OFDM scheme according to the disclosed invention. In S510, RGB video is received. In S520, RGB is converted into a Y-Cr-Cb video data stream. Thus, in one embodiment of the disclosed invention, Y-Cr-Cb video is provided and the conversion described in connection with S510 and S520 is not necessarily required. In S530, for example, decorrelation conversion is performed on each of a plurality of pixel blocks such as 8 × 8 pixels of each Y-Cr-Cb element of the video. Multiple coefficients are generated as a result of each block, for example 64 coefficients in the case of 8 × 8 pixels. Optionally, at S540, for each Y-Cr-Cb element, the number of transmitted coefficients may be selected. One skilled in the art will understand that in this case, some compaction is performed. However, when compaction is performed, compaction is performed to a low coefficient value.

 S550 through S570 provide a more detailed description of the mapping process described in connection with FIGS. 1, 2 and 4 above. In S550, DC and DC approximate range coefficients are processed. Typically, these magnitudes are much larger than the remaining coefficient magnitudes, i.e., the most significant bits (MSBs) are elements of the information to be transmitted, thus they form a coarse display. Therefore, the MSBs of these coefficients are mapped separately and are different from each least significant bit (LSB) referred to as the DC coefficient quantization error. For example, if the coefficient is represented by 11 bits, the 3 MSBs are separated from the remaining bits as a coarse representation and transmitted as their symbols. In one embodiment, this loss of information is so important to the quality of the reconstructed image that the coarse display is repeated with several symbols in order to ensure a proper and coarse display. In particular, a coarse display is transmitted as described in the description related to FIG. 2 above. In another embodiment, error collection codes are used to ensure robust reception of these bits. In addition, error collection is a non-uniform error described in detail in US Provisional Patent Application No. 60 / 752,155 entitled “An Apparatus and Method for Unequal Error Protection of Wireless Video Transmission” granted to co-applicants. Collection, which is incorporated herein by reference to useful information it has. In a further embodiment, the more important coefficients are represented by more MSBs relative to other coefficients represented by fewer MSBs. For example, an LSB of 8 with an 11-bit value (as described above) and an LSB of a DC element having a magnitude explained by the LSB such as the remainder of the high frequency coefficient reconstructs the finer representation of the coefficients, and As described further with reference to FIG. 2, it may be mapped as described in connection with S560 and S570. Each pair of fine representation values may be represented as complex real and imaginary elements that make up the symbols of the OFDM scheme. Thus, if there are 230 available symbols for transmission in a given time slot, up to 460 pairs of complex real and imaginary parts can be transmitted. However, 60 symbols are used to transmit a coarse value as described above. In S580, the symbols are transmitted over the wireless link using an OFDM scheme. The result of using the steps described here provides a very high frame rate of, for example, 45 frames per second or higher, or 0.6 Gbits per second, thus enabling high quality HDTV transmissions. The video is not substantially compressed.

  The distinction between quantized values, referred to as coarse displays, and quantization errors, referred to as fine displays, is generalized as follows to account for DC and other important transform coefficients. These coefficients can be sent through a quantizer that can take several values, ie M = 2 ^ n. The quantizer value specification represented by n bits plays the role of the MSB or the coarse value described above, while the quantization error or fine indication, which is the value represented by the original value-quantizer, represents the LSB described above. To play a role.

  One embodiment of the disclosed invention uses a pilot. Typically, the pilot is transmitted as a well-known signal, preferably a QPSK constellation value, of the bits of the OFDM symbol. These pilots are used in standard modems for synchronization, frequency, phase correction, etc. alone or together with other pilots. Pilots are useful for channel estimation and equalization. In standard digital modems, the latter, which have digital information values utilized via decision feedback, are robust channels because these values are well known to those skilled in the art after decoding. (Robust channel) Enables estimation and tracking. In the invention disclosed herein, analog data transmitted in the manner described in more detail above is prevented from using decision feedback. Thus, in this embodiment of the disclosed invention, additional pilots are transmitted to ensure stable channel estimation and tracking. Here, these pilots may be used for the purpose of transmitting digital data as described above, i.e. coarse values of the transform coefficients are transmitted via these pilots. As additional pilot signals are transmitted, there are many rooms for coarsely displayed data. This results in an improved signal-to-noise ratio (SNR) of the finely displayed data because many parts of the transform coefficients, eg DCT coefficients, are transmitted. Alternatively, relatively important coefficients can be transmitted one or more times, thereby increasing the noise immunity of such important coefficients.

  FIG. 6 shows an exemplary non-limiting block diagram 600 of a system configured to process coding according to the disclosed invention. The baseband modulator is divided into five basic blocks according to the function and working domain of each block. The modulator input consists of four signals, one is a fine data symbol stream, which will be described in detail in connection with the processing of the more important coefficients and the quantization error values of these less important coefficients. Is the result of the converted. The other is, for example, a bitstream representing a coarse portion of the DC values of the Y, Cr and Cb elements, or a coarse portion of other elements described in detail in connection with the MSB of the coefficients described above. These streams are supplied from the video coder 610. In addition, there may be audio and command signals from modem control 670. The signal from modem control 670 consists of a number of commands sent to a receiver, such as the receiver of FIG. 8, and other control signals that control the modulator. In one embodiment of the system 600, the modulator output consists of multiple signals, eg, four signals that carry information to the digital to analog converter 660. This makes it possible to perform the MIMO transmission as described above.

  FIG. 7 is an exemplary and non-limiting block diagram of a bit manipulation unit (BMU) of system 600. The BMU 620 can perform all bit operations on the data bits themselves. There are no quantization errors processed by the BMU 620, and all operations are performed bit by bit. Initially, the audio and coarse representation bitstreams are organized in a predetermined order and the bit composition unit 622 generates a single bitstream. After optional coding by optional coding unit 624, the bits of a single bitstream are mapped to the desired placement by B2S mapper 626 and sent to framer unit 630. Framer unit 630 receives a number of sample streams, a single bit stream and a fine bit stream, and splits them into four sample streams with appropriate headers, pilots, etc. The two different data streams contain pilots and may optionally contain other data as deemed necessary. In the MIMO means, the stream is divided into a plurality of streams, for example four streams, which is for a MIMO transmitter. A frequency domain unit (FDU) 640 obtains input from the framer 530. Framer 630 generates a symbol stream, where each symbol is a complex number according to the disclosed invention as described above indicating the position in two-dimensional space. The framer 630 also includes an inverse fast Fourier transform (IFFT) operation, and the resulting signal is supplied to a time domain unit (TDU) 650 and the signal is converted by a digital to analog converter (DAC) 660. The specific sound signal is shaped before being converted to an analog signal.

  In one embodiment of the disclosed invention, the DAC 660 may operate at a sampling rate of 40 MHz, or a higher frequency, such as 80 or 160 MHz. The desired number of bits includes about 6 dB quantization noise per bit, peak-to-average (PAR) of a signal of 14 dB or less, a desired bit error rate (BER) of 22 dB or less, and a symbol SNR for a given arrangement, and It is approximated using a safety margin condition of 6 dB or less. However, overall, at least 7 bits need to be on the safe side, and due to limitations of the prior art, a 10-bit or 12-bit DAC is used without limiting the generality of the disclosed invention. Is recommended.

  FIG. 8 illustrates an exemplary non-limiting receiver 800 configured to receive signals transmitted in accordance with the disclosed invention. Modulator 810 is configured to receive symbols transmitted in accordance with the disclosed invention, such as, for example, OFDM symbols. For example, reception is performed by means for receiving wireless transmissions received from a plurality of antennas 815. Typically, in a MIMO system, the receiver comprises at least one or more antennas that exceed the number of channels or antennas used by the transmitter. A demodulator 810 receives the signal from antenna 815 processed by unit 812 according to the principles of linear filtering theory and separates the received stream into a coarse stream and a fine stream, respectively. Alternatively, the coarse data is directly decoded by MIMO decoding means such as sphere decoding, while the fine data is processed by linear filtering theory. The fine stream is processed by a decompanding unit 814 that linearizes the received data and generates fine stream data. The coarse data stream is processed by a digital demodulator 816 that operates with standard digital modulation techniques and produces a coarse data stream. The demodulator provides a coarse and fine stream of detected OFDM symbols that are converted to coefficients by unit 820 by reconstructing them. In particular, quantization error information is added to each coarse value that reconstructs DC and approximate DC coefficients. Other fine values constitute high frequency coefficients. Here, the reconstructed coefficients are supplied to an inverse transform unit 830 that generates Y, Cr and Cb elements for video transmission. The color conversion unit 840 further converts the luminance and chrominance inputs to standard RGB output as needed. For purposes of clarity, elements such as, but not limited to, decision feedback equalization, timing and carrier tracking, and other elements that overcome channel distortion and allow accurate reception are not shown. However, these are part of an operable OFDM receiver and are well known in the art and are therefore considered as part of the receiver. The receiver 800 may further be able to receive pilot signals and interpret them as data containing them. Such capabilities are described in detail in US Provisional Patent Application No. 60 / 758,060 entitled "Use of Pilot Symbols for Data Transmission in Uncompressed, Wireless Transmission of Video" This is incorporated herein by reference for useful information it has.

  The present invention has been described herein by reference to several embodiments, including preferred embodiments, but is not limited to other applications, including but not limited to communication according to the disclosed invention over a wire medium. Those skilled in the art will appreciate that the examples can be substituted for the examples described herein without departing from the spirit and scope of the invention. The present invention may be further implemented with hardware, software, firmware, or a combination thereof. Accordingly, the invention should be limited only by the claims. One embodiment may comprise a computer software product that, when executed, comprises a plurality of instructions comprising the invention disclosed herein.

FIG. 1 is a block diagram of a coding system according to the present invention. FIG. 2 is a schematic diagram illustrating an 8 × 8 pixel decorrelation transform, a group of coefficients, and a mapping to a course and excellent symbol display according to the present invention. FIG. 3 is a table showing the number of coefficients selected from each of Y, Cr and Cb after the 8 × 8 pixel conversion according to the embodiment of the present invention. FIG. 4 is a flowchart illustrating the principle of the disclosed invention. FIG. 5 is a flowchart illustrating processing of HDTV video for wireless transmission using the OFDM scheme according to the present invention. FIG. 6 is a detailed block diagram of a coding system according to the disclosed invention. FIG. 7 is a block diagram of a bit manipulation block of a coding system according to the disclosed invention. FIG. 8 is a block diagram of a receiver capable of receiving a transmitted video stream according to the disclosed invention.

Claims (97)

A device for transmitting a high-definition video signal that is substantially uncompressed with almost no delay,
Means for receiving an uncompressed video signal element;
Means for performing a decorrelation transform on the uncompressed video signal elements to provide a plurality of coefficients;
Means for mapping each of said coefficients to a transmitted symbol.   The apparatus of claim 1, wherein the transmission comprises a wireless transmission. The apparatus of claim 1 further comprises:
An apparatus having means for extracting a part of the conversion coefficient.   4. The apparatus according to claim 3, wherein the means for extracting a part of the conversion coefficient further comprises means for extracting the conversion coefficient without depending on a preceding frame of the high-resolution video signal.   The apparatus of claim 1, wherein the transmission of the high resolution video signal includes an unbuffered transmission.   The apparatus of claim 1, wherein the decorrelation transform includes a discrete cosine transform (DCT).   The apparatus of claim 1, wherein the transmission symbols comprise an orthogonal frequency division multiplexing (OFDM) transmission scheme. The apparatus of claim 7, further comprising:
An apparatus comprising an OFDM transmitter for transmitting the transform coefficient.   9. The apparatus of claim 8, wherein the OFDM transmitter comprises means for transmitting via a multi-input multi-output (MIMO) link.   2. The apparatus of claim 1, wherein the uncompressed element comprises an element digital form comprising a luminance signal and a color difference signal.   11. The apparatus of claim 10, further comprising means for converting a color space of the high resolution video signal to the elemental digital format.   4. The apparatus of claim 3, wherein the means for extracting a portion of the transform coefficients extracts only a sufficient number of the coefficients to maintain a substantially uncompressed wireless transmission of the high resolution video signal. A device characterized by comprising means.   13. The apparatus according to claim 12, wherein the means for extracting the conversion coefficient leaves 45 to 64 coefficients out of a total of 64 coefficients related to the luminance signal of the video signal of the high resolution latitude. Device to do.   13. The apparatus according to claim 12, wherein the means for extracting a part of the conversion coefficient leaves 12 to 64 coefficients out of a total of 64 coefficients related to the color difference signal of the high-resolution video signal. Equipment. The apparatus of claim 1, wherein the means for mapping is
The apparatus divides the plurality of coefficients into a first group including a coefficient representing a low frequency and a second group including a coefficient representing a high frequency. The apparatus of claim 15, further comprising:
An apparatus comprising quantization means for generating a coarse value from the first group. The apparatus of claim 16, further comprising:
An apparatus comprising means for mapping a coarse value to one of a plurality of symbol placement points. The apparatus of claim 17, further comprising:
Means for generating additional pilot signals in addition to the standard pilot signals;
Means for utilizing at least one of said additional pilot signals to transmit one of said coarse values. The apparatus of claim 15, further comprising:
Means for generating a quantization error value from the first group;
Means for representing the quantization error value and the coefficient representing a high frequency as a plurality of complex numbers of symbols, wherein each complex value is a first value from the quantization error value and the coefficient representing a high frequency. Comprising both a value and a second value, wherein the first value is represented as a real part of a complex value, and the second value is represented as an imaginary part of the complex value. Device to do.   The apparatus of claim 1, wherein the high resolution video signal has a transmission rate of at least 45 frames per second.   2. The apparatus of claim 1, wherein the high resolution video signal has an uncompressed data rate of 100 Mbps or more per second.   The apparatus of claim 1, wherein the high resolution video signal has an uncompressed data rate greater than 0.6 Gbits per second.   The apparatus of claim 1, wherein said means for performing decorrelation comprises means for processing pixel blocks of said high resolution video signal.   24. The apparatus of claim 23, wherein the pixel block comprises an 8 pixel by 8 pixel rectangle.   24. The apparatus of claim 23, wherein the means for performing decorrelation comprises means for processing a plurality of pixel blocks of the high resolution video signal. The apparatus of claim 1, comprising:
An apparatus comprising means for performing said decorrelation transformation on all pixels of a frame of a video signal. The apparatus of claim 1 further comprises:
An apparatus for performing at least one non-linear transformation of the coefficients.   29. The apparatus of claim 28, comprising means for performing a nonlinear transformation prior to the nonlinear transformation, comprising means for adjusting the power of the nonlinear function of the coefficient to the power of the coefficient. apparatus. A method of transmitting a high-resolution video signal that is not substantially compressed, the method comprising:
Receiving uncompressed elements of the high resolution video signal;
Decorrelating the uncompressed video signal elements to provide a plurality of coefficients;
Mapping each of the coefficients to a transmitted symbol. 30. The method of claim 29 further comprising:
Transmitting the transmission symbol via a wireless link and a wire link.   30. The method of claim 29, wherein the step of extracting a portion of the coefficients is performed independent of a preceding frame of the high resolution video signal.   30. The method of claim 29, wherein the decorrelation transform comprises a discrete cosine transform (DCT).   30. The method of claim 29, wherein the transmission symbols comprise symbols of an orthogonal frequency division multiplexing (OFDM) transmission scheme. 30. The method of claim 29 further comprising:
A method comprising the step of converting the color space of the high resolution video signal to an elemental digital format. 30. The method of claim 29 further comprising:
A method comprising the step of transmitting a symbol stream by an OFDM transmitter. 36. The method of claim 35, further comprising the step of transmitting the symbol stream.
A method comprising transmitting over a multi-input multi-output (MIMO) link. 30. The method of claim 29 further comprising:
A method comprising: extracting a portion of the transform coefficient. 38. The method of claim 37, further comprising extracting a portion of the transform coefficient.
Extracting only a sufficient number of the coefficients to maintain a substantially uncompressed wireless transmission of the high-resolution video signal. The method of claim 38, further comprising the step of retrieving the transform coefficient.
A method comprising leaving 45 to 64 coefficients out of a total of 64 coefficients related to the luminance signal of the high resolution latitude video signal. 40. The method of claim 38, wherein the step of retrieving the transform coefficient comprises:
Leaving 12 to 64 coefficients out of a total of 64 coefficients associated with the color difference signal of the high resolution video signal. 30. The method of claim 29, wherein the mapping step further comprises:
Dividing the plurality of coefficients into a first group including coefficients representing low frequencies and a second group including coefficients representing high frequencies. 42. The method of claim 41, wherein the method of mapping further comprises:
A method comprising the step of generating a coarse value by quantizing the value of the first group. The method of claim 42 further comprising:
A method comprising mapping a coarse value to one of a plurality of symbol placement points. The method of claim 43 further comprising:
Generating an additional pilot signal in addition to the standard pilot signal;
Utilizing at least one of the additional pilot signals to transmit one of the coarse values. The method of claim 42 further comprising:
Generating a quantization error value from the first group;
Representing the quantization error value and the coefficient representing a high frequency as a plurality of complex numbers of symbols, each complex value having a first value from the quantization error value and the coefficient representing a high frequency. Comprising both a value and a second value, wherein the first value is represented as a real part of a complex value, and the second value is represented as an imaginary part of the complex value. how to. 30. The method of claim 29, further comprising the step of performing the decorrelation transformation.
A method comprising processing a pixel block of the high resolution video signal.   48. The method of claim 46, wherein the pixel block comprises an 8 pixel by 8 pixel rectangle.   30. The method of claim 29, wherein the uncompressed element comprises an element digital form comprising a luminance signal and a color difference signal.   30. The method of claim 29, wherein performing the decorrelation transform performs the transform on all pixels of a frame of a video signal. 30. The method of claim 29 further comprising:
A method comprising performing at least one non-linear transformation of the coefficients.   51. The method of claim 50, wherein performing the non-linear transformation comprises adjusting the power of a non-linear function of a coefficient to the power of the coefficient prior to the non-linear transformation. . A system for transmitting and receiving a high-resolution video signal that is substantially uncompressed, comprising:
A transmitter,
Means for receiving uncompressed elements of the high resolution video signal;
Means for performing a decorrelation transform on the uncompressed elements to provide a plurality of coefficients;
Means for mapping each remaining coefficient to a transmitted symbol;
A transmitter comprising means for transmitting a stream of symbols having at least said transmission symbols over a communication link;
A receiver that receives the stream of symbols from the transmitter via a communication link and regenerates the plurality of coefficients from the stream of symbols, wherein the transmitter and the receiver are configured to receive the high-resolution video signal. A system characterized by providing transmission without delay.   53. The system of claim 52, wherein the communication link comprises any wireless link and wire link. 53. The system of claim 52 further comprising:
A system comprising means for extracting a part of the transform coefficient, wherein the means for extracting a part of the transform coefficient extracts it without depending on a preceding frame of the high-resolution video signal.   55. The system of claim 54, wherein the transmitter and the receiver comprise means for unbuffered transmission and reception of the high resolution video signal.   53. The system of claim 52, wherein the transmission symbols comprise symbols of an orthogonal frequency division multiplexing (OFDM) transmission scheme. 53. The system of claim 52, wherein the transmitter is
A system comprising means for transmitting data comprising said high resolution video signal via a multi-input multi-output (MIMO) link. 58. The system of claim 57, wherein the receiver further comprises:
A system comprising means for receiving data comprising said high resolution video signal from a multi-input multi-output link. 53. The system of claim 52 further comprising:
Means for generating additional pilots;
Means for receiving said additional pilot,
The system characterized in that at least one of the additional pilots is a most significant bit of a digital value that prevents the decorrelation transformation from being performed. 53. The system of claim 52, wherein the means for mapping is
A system comprising means for dividing the plurality of coefficients into a first group comprising coefficients representing low frequencies and a second group comprising high frequencies. The system of claim 60 further comprising:
A system comprising quantization means for generating a coarse value from the first group. The system of claim 61 further comprising:
A system comprising means for mapping a coarse value to one of a plurality of symbol placement points. The system of claim 62 further comprising:
Means for generating additional pilot signals in addition to the standard pilot signals;
Means for utilizing at least one of the additional pilot signals to transmit one of the coarse values. The system of claim 60 further comprising:
Means for generating a quantization error value from the first group;
Means for representing the quantization error value and the coefficient representing a high frequency as a plurality of complex numbers of symbols, wherein each complex value is a first value from the quantization error value and the coefficient representing a high frequency. Comprising both a value and a second value, wherein the first value is represented as a real part of a complex value, and the second value is represented as an imaginary part of the complex value. System. The symbol-based transmission method is at least
Receiving an input stream of data samples;
Generating a first symbol in a fine display arrangement based on sample values corresponding to at least a portion of the input stream.   66. The method of claim 65, wherein the input stream comprises a more important part and a less important part. The method of claim 66 further comprising:
Generating a second symbol in a coarse display arrangement based on at least a portion of the input stream. 68. The method of claim 67, wherein generating the second symbol comprises:
Quantizing a significant portion of the input stream to a coarse value;
Mapping a coarse value to the second symbol. The method of claim 68 further comprising:
A method comprising: mapping a quantization error value obtained from quantizing a highly significant portion of the input stream to the first symbol. The method of claim 66 further comprising:
A method comprising mapping a less important part of the input stream to the first symbol. 66. The method of claim 65, wherein generating the first symbol comprises:
Mapping the first value to a complex real value;
Mapping a second value to the complex imaginary value. The method of claim 71 further comprising:
Providing the contents of the first value from at least a portion of the input stream and a quantization error value of a quantized significant portion of the input stream;
Providing the contents of the second value from at least a portion of the input stream and one of the quantization error values of the quantized significant portion of the input stream. A method characterized by. The method of claim 71 further comprising:
Transmitting the first symbol on a low transmission quality subchannel. The method of claim 65 further comprising:
A method comprising performing a non-linear mapping on at least a portion of the input stream. The method of claim 65 further comprising:
Adjusting the power obtained by the non-linear mapping of at least a portion of the input stream to the power of the portion of the input stream preceding the non-linear mapping. The method of claim 65 further comprising:
Transmitting the symbol over either a wireless link or a wire link.   77. The method of claim 76, wherein the transmission occurs over a wireless link comprising a multi-input multi-output link.   66. The method of claim 65, wherein the symbols comprise OFDM symbols.   68. The method of claim 65, further comprising generating a plurality of pilot symbols. The method of claim 79 further comprising:
A method comprising generating at least one additional pilot symbol to transmit a value of high importance.   66. The method of claim 65, wherein the input stream corresponds to a video frame transform coefficient.   82. The method of claim 81, wherein the coefficient corresponds to a decorrelation transform of either a luminance component of a video frame or a chrominance component of the video frame.   82. The method of claim 81, wherein the less important portion corresponds to either a coefficient of a video frame or a quantization error value of a coefficient corresponding to a high frequency. An orthogonal frequency division multiplexing (OFDM) communication system comprising:
Means for receiving data values ranked by importance;
Means for mapping data values of less important data values as imaginary and real values of fine OFDM symbol placement;
Means for mapping OFDM symbols to subchannels of a communication channel. The OFDM communication system of claim 84 further comprising:
Means for mapping a quantized value of a coarse representation of a highly important data value to an OFDM symbol;
Means for mapping the quantization error value of the quantized value of the high importance data as a pair of imaginary and real values of fine OFDM symbol constellation. The OFDM communication system of claim 84 further comprising:
A system comprising means for transmitting OFDM symbols corresponding to low ranking data over a low quality subchannel. The OFDM communication system of claim 84 further comprising:
A system comprising means for transmitting the OFDM symbol over either a wireless link or a wire link.   88. The OFDM communication system of claim 87, wherein the wireless link comprises a multi-input multi-output link. The OFDM communication system of claim 84 further comprising:
A system comprising means for generating at least one additional pilot symbol for transmission of the coarse indication. The OFDM communication system of claim 84 further comprising:
A system comprising means for compressing and decompressing at least a portion of the data value. The OFDM communication system of claim 84 further comprising:
A system comprising non-linear mapping means for at least a portion of the data value. The OFDM communication system of claim 84 further comprising:
A system comprising means for transmitting fine OFDM symbols on a low transmission quality subchannel.   85. The OFDM communication system according to claim 84, wherein the data value corresponds to a coefficient of video frame conversion.   94. The OFDM communication system according to claim 93, wherein the coefficient corresponds to a decorrelation between a luminance element of a video frame and a chrominance element of the video frame.   94. The OFDM communication system according to claim 93, wherein the less important part corresponds to either a coefficient of a video frame or a quantization error value of a coefficient corresponding to a high frequency.   94. The OFDM communication system according to claim 93, wherein the less important data values of the coefficients of video frame conversion correspond to DC and approximate DC coefficients.   94. The OFDM communication system according to claim 92, wherein non-important data values such as video frame conversion coefficients correspond to high frequency coefficients.

Images