Multiplex communications – Communication over free space – Repeater
Reexamination Certificate
1998-09-03
2002-09-03
Patel, Ajit (Department: 2662)
Multiplex communications
Communication over free space
Repeater
C370S342000, C455S522000
Reexamination Certificate
active
06445686
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to wireless communications, and more particularly, to a method and apparatus for improving the quality of speech signals transmitted over a wireless communications system.
BACKGROUND OF THE INVENTION
In order to maximize use of the limited available bandwidth for wireless communication services, various compression techniques and multiplexing techniques are used on the link between a base station, which is connected to a wireline telecommunications network, and a plurality of different users at mobile units each of whom are simultaneously communicating, via that base station, with other users who may be connected to either a wired or wireless network. Time Division Multiple Access (TDMA) and Code Division Multiple Access (CDMA) are examples of two well known multiplexing schemes used in wireless cellular and PCS systems. In TDMA, a time frame is subdivided into a plurality of time slots and a user's mobile unit communicates in burst transmissions over a specific time slot at uplink and downlink carrier frequencies associated with the base station, which time slot assignment is made when a call commences. A receiver at a base station time-division multiplexes the plural digital speech signals directed to the plural mobile units into a downlink signal which is transmitted at a downlink carrier frequency to the mobile units. The receiver in each mobile unit then recovers the digital speech signal directed to itself in accordance with the time slot assigned to that mobile unit. Similarly, the uplink signal transmitted by each mobile unit is transmitted in an assigned time-slot, which time-slot assignment is used by the base station to appropriately recover the digital speech signal transmitted by the mobile unit. In CDMA, rather than assigning a mobile unit a specific time slot, a mobile unit is assigned a specific encryption code which is used to spread the spectrum of the coded speech signal over the channel. The plural mobile units that are simultaneously communicating with the base station are each assigned a different encryption/decryption code. The codes associated with each mobile unit are then used by the base station to multiplex the plural digital input speech signals into a downlink signal which is transmitted to all the mobile units. The mobile unit set then uses its associated decryption code to recover the particular digital speech signal directed to it from the multiplexed downlink signal transmitted by the base station. Uplink communication from the plural mobile units functions in a parallel manner.
In order to reduce the number of bits representing the coded digital speech signal, and therefore improve the efficiency of use of the available frequency bandwidth, speech compression techniques are used in wireless communication. The analog speech signal, which is normally sampled at, for example, an 8 kHz rate to produce a PCM bit stream, is subdivided into frames and compressed using an appropriate coding algorithm. In TDMA systems, a fixed rate vocoder, such as an ACELP (Algebraic Code Excited Linear Predictive) or VCELP (Vector-Sum CELP) coder, is used to compress the PCM samples. In CDMA systems, a variable rate CELP algorithm is used. Specifically, for CDMA, a speech encoder produces a variable rate output based on the speech activity of the input speech signal. During active speech periods, the speech encoder produces full rate 20 ms frames. During the silent periods, the speech encoder produces ⅛ rate frames. During transition periods between the talking periods and silent periods, the speech encoder produces ½ rate and ¼ rate frames. During these sub-rate frames, power consumption is reduced to a lower level than during full rate frames advantageously reducing the overall power output.
In the CDMA transmitter at either the mobile unit or at the base station, the PCM speech samples are broken down into 20 ms frames. The speech encoder uses an analysis-by-synthesis method to optimally determine the parameters for a given PCM speech frame input. For every 20 ms PCM speech frame input, the speech encoder produces a set of output parameters that represent the encoded frame. The speech encoder determines that set as a set of input parameters for an internal-to-the-encoder decoder, which minimizes the perceptual difference between the synthesized speech, which is output from the internal decoder, and the original input speech. An encoded full rate frame includes the following parameters: the linear predicative coding (LPC) parameters; pitch lag (L) and pitch gain (b) parameters; and codebook gain (G) and codebook index (I) parameters. At the receiving end, at the mobile unit or at the base station, the speech decoder receives each incoming encoded frame and converts each frame back into a sequence of PCM speech samples using the encoded frame. (see, e.g., TIA [Telecommunications Industry Association] IS-733-High-Rate [13 k] bits per second Speech Service Option).
As a consequence of the susceptibility of wireless links to interference and other inherent atmospheric conditions, a transmitted frame may not reach the receiver at either the base station or mobile unit set or may be severely corrupted by noise or interference. When a frame is “lost” or is so corrupted with noise or interference as to be undecodable, it is marked by the receiver as being erased and no encoded parameters are supplied to the speech decoder. In order to minimize the perceptual effect of such an erased frame, a Frame Masking Algorithm is used to estimate the PCM samples for the erased frame using an extrapolation of the data from a previous frame. Thus, the speech decoder uses the previous values of the aforedescribed frame parameters to determine the current values of the erased frame. More specifically, for CDMA 13 k systems, the current values of the linear predictive coding (LPC) parameters are determined by decaying the LPC parameters of the previous frame, where the decaying coefficient is a function of the number of consecutive erasure frames. The current value of the pitch gain lag (L) is repeated from the previous frame; the current value of the pitch gain (b) is determined from the pitch gain of the last frame; the current value of the codebook gain (G) is determined by subtracting an appropriate integer from the previous value of G; and the codebook index (I) is determined randomly. A problem arises, however, when the previous frame to an erased frame is less than a full rate frame. When the previous frame is, for example, a ⅛ rate frame, the resultant perceptive speech quality obtained by extrapolating parameters from the ⅛ rate frame is poor.
In the prior art, the power of the signals transmitted on the downlink from the base station to the mobile unit set and on the uplink from the mobile unit to the base station is controlled to minimize power while maintaining an acceptable frame error rate. In particular, on the downlink, a power control algorithm makes a decision to increase or decrease the base station transmit power based on information provided by the mobile unit on the uplink. The mobile unit monitors the downlink and compiles statistics about the downlink frame error rate. This information is then conveyed back to the base station on the uplink to enable the base station to control its transmit power to maintain the desired downlink error statistics. On the uplink, the mobile unit controls its transmit power in direct response to power control order messages sent by the base station to the mobile unit on the forward link. The uplink power control algorithm determines the required direction of uplink transmit upward or downward power change and the magnitude of the change based on the history of the received frames to date. The history of the frames received on the uplink includes good frames, corrupt frames, and frame rate information.
Disadvantageously, such prior art power control schemes increase the power transmitted by the base station and the mobile unit onl
Hoffbeck Joseph P.
Meyers Martin Howard
Premji Al-Nasir A.
Tarraf Ahmed A.
Tracy Kevin G.
Gurey Stephen M.
Lucent Technologies - Inc.
Patel Ajit
Tsegaye Saba
LandOfFree
Method and apparatus for improving the quality of speech... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for improving the quality of speech..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for improving the quality of speech... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2829199