Patent Application
20070177670
1. Field of the Invention
The invention relates to the use of pilot symbols in the transmission of uncompressed video over a wireless link. More specifically, the invention relates to use of pilots to transmit data symbols in video transmission where direct mapping of image transform coefficients to transmission symbols is performed.
2. Discussion of the Prior Art
In many houses, television and/or video signals are received through cable or satellite links at a set-top box that is located at a fixed point in the house. In many cases, it is desired to place a screen at a point a distance from the set-top box by a few meters. This trend is becoming more common as flat-screen using plasma or liquid crystal display (LCD) televisions are increasingly hung on a wall. Connection of the screen to the set-top box through cables is generally undesired for aesthetic reasons and/or installation convenience. Thus, wireless transmission of the video signals from the set-top box to the screen is preferred. Similarly, it may be desired to place a computer, game controller, VCR, DVD, or other video source that generates images to be displayed on a screen a distance from the screen.
Generally, the data received at the set-top box are compressed in accordance, for example, with the motion picture expert group (MPEG) format and are decompressed by the set-top box to a high quality raw video signal. The raw video signal may be in an analog format or a digital format, such as the digital video interface (DVI) format or the high definition multimedia interface (HDMI) format. These digital formats generally have a high definition television (HDTV) data rate of up to about 1.5 Giga bits per second (Gbps).
Wireless short range transmission in the home can be accomplished over the unlicensed bands around 2.4 GHz or around 5 GHz, e.g. in the U.S in the 5.15-5.85 GHz band. These bands are currently used by wireless local area networks (WLAN), where the 802.11 WiFi standard allows maximal data rates of 11 Mbps (802.11b), or 54 Mbps for 20 MHz bandwidth using the 802.11g/802.11a standards. With the emerging Multi-Input Multi-Output technology the data rate of the 802.11n standard is increased to around 200 Mbps. Another alternative is to use Ultra Wide Band (UWB), which claims to provide 100-400 Mbps.
Because the available data rate is lower than the 1.5 Gbps needed for uncompressed HDTV video, the video generally must be recompressed for wireless transmission, when desired. Known strong video compression methods, e.g. those having a compression factor of above 1:30, require very complex hardware to implement the compression. This is generally not practical for home applications. These compression methods generally transform the image into a different domain by using, for example, wavelet, discrete cosine transform (DCT), or Fourier transforms, and then perform the compression in that domain. In PCT application IL/2004/000779, Wireless Transmission of High Quality Video, assigned to common assignee and incorporated herein in its entirety by this reference thereto, there is discussed a method of transmitting video images. The method includes providing high definition video, compressing the video using an image domain compression method in which each pixel is coded based on a vicinity of the pixel, and transmitting the compressed video over a fading transmission channel.
U.S. patent publication 2003/002582 by Obrador describes wireless transmission of images which are encoded using joint source channel coding (JSCC). The transmitted images are decomposed into a plurality of sub-bands of different frequencies. Image and corresponding boundary coefficients with a lowest resolution are sent first, and then image and boundary coefficients with a higher resolution are transmitted. An exemplary JSCC applies channel encoding techniques to the source coded coefficients, providing more protection to more important, i.e. low frequency, coefficients and less protection to less important, i.e. high frequency, coefficients.
In digital transmission methods, signals are transmitted in the form of symbols. Each symbol can have one of a predetermined number of possible values. The set of possible values of each symbol is referred to as a constellation and each possible value is referred to as a bin. In two dimensional constellations, the distance between neighboring bins affects the immunity of the symbols to noise. The noise causes reception of the symbol in a bin that may not be the intended bin. If, due to the noise, the symbol is closer to a different bin than intended, the symbol may be interpreted incorrectly. See Ramstad, The Marriage of Subband Coding and OFDM Transmission, Norwegian University of Science and Technology (July 2003).
In U.S. patent application serial nos. 2004/0196920 and 2004/0196404 by Loheit et al. another scheme is proposed for the transmission of HDTV over a wireless link. The discussed scheme transmits and receives an uncompressed HDTV signal over a wireless RF link which includes a clock that provides a clock signal which is synchronized to the uncompressed HDTV signal. This scheme also includes a data regeneration module that is connected to the clock, and which provides a stream of regenerated data from the uncompressed HDTV signal. A demultiplexer demultiplexes the stream of regenerated data using the clock signal into an I data stream and a Q data stream. A modulator connected to the demultiplexer modulates a carrier with the I data stream and the Q data stream. According to Loheit et al., the RF links operate at a variety of frequency bands from 18 GHz up to 110 GHz, hence requiring sophisticated and more expensive transmitters and receivers.
To provide better reception, pilot symbols are used in OFDM transmission. The pilot symbols are used for the purpose of enabling synchronization of the reception to the channel characteristics, thereby enabling a better and more accurate reception of the transmitted data. This is of particular importance in systems where retransmission of data is not possible, for example in the case of a bandwidth limited channel, such as is typically found where there is a need to transmit HDTV signals over a wireless link. However, the use of the pilot signals that are known to both the transmitter and the receiver reduces the effective bandwidth because fewer symbols are made available for transmission of actual data. A training session may therefore be used periodically to attempt to overcome this limitation. Nonetheless, this is still a restriction on the performance of the channel. Moreover, this scheme does not overcome drift in channel characteristics in real-time.
In view of a variety of limitations of the prior art, it would be therefore advantageous to provide a solution that enables the reliable wireless transmission of an HDTV stream, while avoiding the need to dedicate a portion of the available bandwidth to the transmission of known data solely for the purpose of pilot symbols.
The uncompressed wireless transmission of video, as with many other wireless applications, requires constant knowledge of channel characteristics at the receiver end. To estimate the channel and track its changes, pilots containing known data are sent in various parts of the used bandwidth. The use of such pilots reduces the effective bandwidth available for data transmission. Due to the relative high immunity to introduced interference of certain transmission modes, such as QPSK and QAM, pilots can be modulated by digital data components. At the receiver, pilots are demodulated and used for a decision-directed circuit to determine the characteristics of the transmission channel. The additional bandwidth allows a higher data rate which may be such used for various purposes as diversity, coding, etc. Such use of pilot signals is of particular advantage in the wireless transmission of the DC and near DC components of essentially uncompressed video.
The uncompressed wireless transmission of video, as with many other wireless applications, requires the constant knowledge of the channel characteristics at the receiver. To estimate the channel and track its changes, pilots containing known data are sent in various parts of the used bandwidth. The use of such pilots reduces the effective bandwidth available for data transmission. Due to the relative high immunity to introduced interference of certain transmission modes, such as QPSK and QAM, pilots can be modulated by digital data components. At the receiver, pilots are demodulated and used by a decision-directed circuit to determine the characteristics of the transmission channel. The additional bandwidth allows higher a data rate that can be used for various purposes, such as diversity, coding, etc. Such use of pilot signals is of particular advantage in the wireless transmission of the DC and near DC components of essentially uncompressed video. See, for example, U.S. patent application entitled: Apparatus and Method for Uncompressed, Wireless Transmission of Video, Ser. No. 11/551,641, is incorporated herein in its entirety by this reference thereto.
The invention disclosed herein is better understood with respect to a video transmission system enabling the mapping the coefficients of a block of pixels after a de-correlating transformation, or a portion thereof, directly into the transmission symbols. Preferably, a discrete cosine transform (DCT) is performed on a block of pixels of each of the Y, Cr, and Cb components of a video signal. The Y component provides the luminance of the pixel, while the Cr and Cb components provide the color difference information. In a preferred embodiment of the invention, only a portion of the coefficients are used for transmission purposes, avoiding the very high frequency coefficients while keeping the lower frequency coefficients. Significantly, more of the Y related coefficients are preserved for wireless transmission purposes than those for the other two components. For example, a ratio of at least three coefficients of the Y component may be used for each of the Cr and Cb components, e.g. a ratio of 3:1:1. DC coefficients, or proximate coefficients having a larger value, are also represented in a digital manner, i.e. part of the DC value is represented as one of a plurality of constellation points of a symbol. The higher frequency coefficients and, in addition, the quantization errors of the DC and the proximate components whose main part is presented digitally are grouped in pairs, positioning each pair in a symbol as the real and imaginary values of the complex number. Optionally, a possibly non-linear transformation of these values is represented as a complex number of that mapped to an OFDM component.
Following is a detailed description of the operation of the invention. While the invention is described with respect to particular embodiments and respective figures, such are not intended to limit the scope of the invention and are provided for purposes of example only.
In one embodiment of the invention, the block 120 performs a DCT on the blocks of pixels. A block of pixels may contain 64 pixels arranged in an 8-by-8 format, as shown in to
The output of transform unit 120 is a series of coefficients which are fed to a mapper 130. The mapper 130 selects those coefficients from each of the Y-Cr-Cb components which are to be transferred over the wireless link. The mapper 130 also maps the coefficients to be sent to transmission symbols, for example, the symbols of an orthogonal frequency division multiplexing (OFDM) scheme. This process is described in more detail with respect to
In one embodiment of the invention, a modified OFDM transmitter 140 is connected to a plurality of antennas for the purpose of supporting a multi-input, multi-output (MIMO) transmission scheme, thereby increasing bandwidth and reliability of the transmission. A person skilled in the art would further appreciate that a receiver adapted to receive the wireless signal comprising the symbols transmitted in accordance with the invention must be capable of detecting the digital and analog symbols, reconstructing the respective coefficients, and performing an inverse transform to reconstruct the Y-Cr-Cb components. However, because there is no frame-to-frame compression, there is no need for frame buffers in the system. Because the mapping and transform are fast and work on small blocks with no need to consider neither wide area correlation in the image nor frame-to-frame correlations, there is practically no delay associated with the operations disclosed herein.
In accordance with the invention, a de-correlating transform, such as a DCT, is performed on blocks of pixels, for example 8-by-8 pixels, on each of the Y-Cr-Cb components of the video. As a result of the transform on a block, for example a block 210 shown in
The coefficient 222-i is mapped to one such constellation point, depending on its specific value. However, it is also likely to have a quantization error, or in other words, a value corresponding to the difference between the original value and the value represented by the digital bits that are mapped to constellation points. This error essentially corresponds to the least significant bits of the coefficient's value. Such a mapping is considered a digital value mapping. The quantization error value may be mapped as part of the symbol 240-i as, for example, the real portion of the complex number constituting the symbol 240-i. The higher frequency coefficients are paired and each pair is mapped as a real portion and an imaginary portion of a complex number. For example, the coefficients 224-i and 224-j may be mapped to the imaginary and real portions of a symbol 240-j. As noted above, a receiver enabled to receive the symbol stream disclosed herein should be able to recompose the coefficients from the values provided. Such a mapping is considered an analog value mapping. It should be noted that the transferred data may be coded or uncoded.
An exemplary reference may be found in
Steps 450 through 470 provide a more detailed description of the mapping process discussed with respect to
In another embodiment an error correction code is used to assure the robust reception of these bits. The LSBs of the DC component, as well as the rest of the coefficients that (as noted above) have an amplitude described by the LSBs, for example 8 LSBs of an 11-bit value, may be mapped, as explained with respect of steps 460 and 470, and as further discussed with reference to
The separation to MSB and LSB in describing the DC and other important transform coefficients can be generalized as follows: These coefficients can pass via a quantizer that can take several values, E−6 M=2ˆn. The specification of the quantizer value, represented by n bits, plays the role of the MSB's above, while the quantization error, i.e. the original value minus the value represented by the quantizer, plays the role of the LSB's above.
One embodiment of the invention makes use of pilots. Commonly, pilots are sent as a priori known signals in some bins of the OFDM symbol, preferably a value from a QPSK constellation. These pilots, alone or in conjunction with other pilots, are used in standard modems for synchronization, frequency, phase corrections, and the like. Pilots can also help in channel estimation and equalization. In standard digital modem, these pilots together with the digital information values, the latter being used via decision feedback because these values are known to those skilled in the art after decoding, allow robust channel estimation and tracking. Referring to
Specifically, in the invention, the analog data sent in the manner discussed in more detail above, makes the use of decision feedback impossible. Therefore, in accordance with this embodiment of the invention additional pilots are sent to ensure stable channel estimation and tracking. These pilots are used for sending the digital data discussed in more detail above, i.e. MSBs of some transform coefficients are sent over these pilots. Because additional pilot signals are sent, there is more room for digital data. This results in an improved signal-to-noise ratio (SNR) on the analog data because even larger portion of the DCT coefficient is now sent digitally. In one embodiment of the invention approximately 30% of the sent data over the wireless channel is the digital portion, as explained above in more detail, and which may be used in accordance with the disclosed method for a decision directed correction of the received symbols using, for example, least mean square (LMS) techniques.
First, three bit streams are arranged in a predefined order and create a single bit stream. After optional coding, the bits of the single bit stream are mapped to the desired constellation and passed to a framer unit 530. The framer unit 530 receives the data as a number of sample streams and organizes it into four sample streams with an appropriate header, pilots, and so on. Two different data streams are padded with pilots and, optionally, with some other data where it may be deemed necessary, and then interleaved.
In a MIMO implementation, the stream is divided into a plurality of streams, for example four streams, one for each of the MIMO transmitters. The frequency domain unit (FDU) 540 gets its inputs from the framer 530. The framer 530 creates a symbol stream, such that each symbol is a complex number, as described hereinabove, that represents a point in the two-dimensional space. The framer 530 also includes an IFFT operation, and the resultant signal is fed to the time domain unit (TDU) 550, where certain shaping of the signal is performed prior to converting the signal to an analog signal in the digital-to-analog converter (DAC) 560.
The DAC 560 may be operative, in one embodiment of the invention, at a sampling rate of 40 MHz, or even higher frequencies, for example 80 or 160 MHz. The desirable number of bits can be approximated using the following assumptions: quantization noise of about 6 dB per bit, peak to average (PAR) of the signal ˜14 dB, symbol SNR for a desired bit error rate (BER) and given constellation ˜22 dB, and a safety margin ˜6 dB. In total, at least seven bits are required, however, to be on the safe side, and according to the limitations of existing technology, it is recommended to use, without limiting the generality of the invention, a 10-bit or even 12-bit DAC.
Although the invention is described herein with reference to several embodiments, including the preferred embodiment, one skilled in the art will readily appreciate that other applications may be substituted for those set forth herein without departing from the spirit and scope of the invention. The invention may be further implemented in hardware, software, or any combination thereof. Accordingly, the invention should only be limited by the following Claims.
This application claims priority from U.S. Provisional Patent Application No. 60/758,060 filed on Jan. 10, 2006 which is incorporated herewith in its entirety by the reference thereto.
| Number | Date | Country | |
|---|---|---|---|
| 60758060 | Jan 2006 | US |
