Pulse or digital communications – Spread spectrum – Direct sequence
Reexamination Certificate
2001-02-08
2002-04-09
Chin, Stephen (Department: 2634)
Pulse or digital communications
Spread spectrum
Direct sequence
C375S147000, C375S260000, C375S343000, C370S335000, C370S336000, C370S441000, C370S442000
Reexamination Certificate
active
06370182
ABSTRACT:
Reference is also made to Weinberg et al application Ser. No. 09/382,202 filed Aug. 23, 1999 and entitled MULTI-BAND, MULTI-FUNCTION, INTEGRATED TRANSCEIVER which is incorporated herein by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates in general to wireless communication receivers. In particular, it relates to the integration of multiple signal types (CDMA, FDMA, CW, etc.), from multiple bands, with each band and signal type potentially containing multiple user channels, and a single receiver processing architecture with multiple antenna elements per band for sequentially acquiring, and simultaneously demodulating these multiple channels, utilizing jointly-optimized advanced signal processing techniques of digital beamforming, Rake multipath combining, and joint detection.
2. Description of the Prior Art
Matched Filtering
A matched-filter is typically employed in a spread-spectrum demodulator to remove the effects of PN-spreading and allow the carrier and modulating information to be recovered. The digital implementation of a matched filter can be expressed as an integrate-and-dump correlation process, which is of relatively modest computational burden during signal tracking and demodulation. However, it is computationally and/or time intensive to acquire such a signal, where many such correlations must be performed to achieve synchronization with the transmitted spreading sequence. For each potential code-phase offset to be searched (which typically number in the thousands), sufficient samples must be correlated to ensure that the integrated SNR is sufficient for detection. Performed one at a time, acquisition could easily take several minutes to achieve in typical applications.
For applications requiring rapid signal acquisition (e.g., seconds), a highly parallel matched-filter structure may be used to search many spreading code offsets simultaneously. Typically, this computationally expensive apparatus would be underutilized once acquisition is completed, during the much less demanding tracking operation. If the same parallel matched filter is also used for tracking purposes, only perhaps three of its numerous correlation branches (perhaps hundreds) are useful in this instance. Alternatively, it may be simpler to use a separate set of early, on-time, and late integrate-andidump correlators to take over once acquisition is complete; in this case, the parallel matched filter would go completely unused during tracking.
In implementations evidenced by the prior art, the matched-filtering solution has generally fallen into one of several classes:
1. Slow acquisition by sequential traversal of the search space using only the hardware required for tracking a signal; dedicated hardware per channel.
2. Rapid acquisition by parallel traversal of the search space using a dedicated parallel matched filter, which is idle or shut down when dedicated tracking hardware takes over; dedicated hardware per channel.
3. Either class 1 or 2, but multi-band and/or multi-channel, using a loosely integrated but disparate collection of individual processing resources.
Beamforming
Beamforming is a form of spatial filtering in which an array of sensor elements are utilized with appropriate signal processing to digitally implement a phased array antenna, for the purpose of shaping the antenna response over time in a space-varying manner (i.e., steering gain in some directions, and attenuation or nulls in other directions). In a radio communications system, a signal arriving at each element of an antenna array will arrive at slightly different times, due to the direction of arrival with respect to the antenna array plane (unless it has normal incidence to the plane, in which case the signal will arrive at all elements simultaneously). A phased array antenna achieves gain in a particular direction by phase-shifting, or time-shifting, the signal from each element, and then summing them in a signal combiner. By choosing the relative phasing of each element appropriately, coherence can be achieved for a particular direction of arrival (DOA), across a particular signal bandwidth.
Digital beamforming is very analogous to this, except that the signal on each antenna element is independently digitized, and the phasing/combining operation performed mathematically on the digital samples. Traditionally, digital beamforming is done on a wideband signal, prior to despreading a CDMA waveform. This forces the computationally intense beamforming to take place at a much higher sampling rate, resulting in more mathematical operations per second, and corresponding increased hardware cost (there are examples addressing this shortcoming in the prior art, such as Hanson et al., where beamforming is performed at baseband to avoid this and other issues).
Furthermore, digital beamforming is traditionally done as a separate process, independent of symbol demodulation, perhaps even as a separate product from the demodulator. In addition to the resulting inability to support advanced demodulation techniques with this architecture, the cost of the beamforming function is greater as a stand-alone function, compared to the incremental cost of adding the capability to a demodulator. The largest cost-component of beamforming is the complex multiplication of each sample for each element with the beamforming weights. When combined with the demodulator, the complex multiply can be absorbed into computation already taking place for extremely low incremental cost due to beamforming (there is, for example, an implementation of beamforming using digital direct synthesis (DDS) functions in the prior art, such as Rudish, et al.). Thus, whether stand-alone beamformers merely point in the direction of the signal of interest, or respond more adaptively to dynamic interference conditions by null-steering, they still lack the ability to be tightly coupled with potential advanced demodulation techniques.
Rake Combining
Rake combining is a method of mitigating the effects of a multipath interference dominated communications channel, as is adaptive equalization. However, in a typical equalizer, the filter time-span must correspond to the multipath delay spread, and therefore tends to be limited to very close-in multipath, spanning perhaps a few symbols. The Rake, however, exploits the properties of CDMA signals (i.e., during despreading, all other codes become uncorrelated, including copies of the desired code delayed by greater than about half a chip, and are reduced to noise across the entire spread bandwidth) that enables each multipath component (offset by more than about half a chip) to be acquired, tracked, and despread in isolation, and then coherently combined. Much like beamforming, this coherent combining results in increased effective antenna aperture and improved SNR, although using only a single antenna element. This divide-and-conquer approach allows the Rake to span an essentially arbitrary multipath delay spread, applying computational resources based linearly on the number of desired despreader branches, or “Fingers”, desired, and not based on the delay spread itself (although acquisition time, and thus dynamic performance, is related to the actual delay spread, as this defines the limits of what must be searched).
In the prior art, Rake combining is typically employed as a dedicated function in a fixed CDMA receiver structure. Resources are designed into the receiver to perform some fixed maximum number of Rake Fingers, and those resources are tied up regardless of whether those Fingers are actually utilized or not. What is needed is a more flexible and generalized receiver architecture, which can task resources on more of a demand basis, and furthermore treat diversity information such as Rake Fingers as simply one of several diversity inputs to be jointly optimized in a common process that yields maximum advantage to each desired user signal.
What is needed is the ability to combine potential spatial processing information with other dimensions of information and diversity, both regarding t
Bierly Scott
Harlacher Marc
Smarrelli Robert
Weinberg Aaron
Chin Stephen
Ha Cac V.
ITT Manufacturing Enterprises Inc.
Zegeer Jim
LandOfFree
Integrated beamforming/rake/mud CDMA receiver architecture does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Integrated beamforming/rake/mud CDMA receiver architecture, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Integrated beamforming/rake/mud CDMA receiver architecture will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2895435