Multiuser communications system

• US 10,284,286 B2
• Filed: 08/21/2014
• Issued: 05/07/2019
• Est. Priority Date: 08/21/2013
• Status: Active Grant

• Current Assignee: Myriota Pty Ltd.
• Original Assignee: Myriota Pty Ltd.
• Inventors: Grant, Alexander James, Lechner, Gottfried, Pollok, Andre, McKilliam, Robert George, Land, Ingmar Rudiger, Haley, David Victor Lawrie, Cowley, William George, Davis, Linda Mary, Barbulescu, Sorin Adrian, Buetefuer, John Lawrence, Lavenant, Marc Pierre Denis
• Primary Examiner(s): Elallam, Ahmed
• Application Number:
  US14/913,339
• Publication Number:
  US 20160204852A1
• Time in Patent Office: 1,720 Days
• CPC Class Codes: H04B 1/7107 Subtractive interference cancellation
• CPC Class Codes: H04B 7/18539 Arrangements for managing radio, resources, i.e. for establishing or releasing a connection
• CPC Class Codes: H04W 72/0446 Resources in time domain, e.g. slots or frames
• CPC Class Codes: H04W 72/541 using the level of interference
• CPC Class Codes: H04W 88/02 Terminal devices

Abstract

A multiuser communication system comprises multiple transmitters and a multiuser receiver that detects multiple transmissions via iterative soft interference cancellation. An initial acquisition module and single user decoder module are also described. The multiuser receiver acquires and subtracts known users in the residual signal before acquiring new users in the residual signal, which is performed iteratively until no new users are detected or a stopping criterion is met. To aid receiver acquisition, the transmitters insert discrete tones into the transmitted signals. These allow the multiuser receiver to obtain initial estimates of the frequency, time, gain, and/or phase offset for each user. To improve the quality of cancellation the receiver refines estimates of gain, time, frequency and phase offsets for each user after each iteration, and calculates time varying SINR estimates for each user. The multiuser receiver may be satellite based, may be a distributed receiver, or process users in parallel.

Claims

1. A method for operation of a multiuser receiver in a slot based wireless communication system to detect multiple transmissions via iterative soft interference cancellation, the method comprising:

a) receiving a signal in a slot;

b) attempting to acquire one or more user signals in a residual signal by searching for a predefined signal inserted in a complex transmit signal transmitted by a user to aid acquisition, and for each new user signal acquired, estimating a set of channel parameters for the acquired new user signal, and adding the acquired new user signal to a set of acquired user signals, wherein the set of channel parameters comprises at least one of a complex gain, a time offset, a frequency offset, a rate of change of frequency offset, or a noise variance;

c) decoding the residual signal, for each acquired user signal in the set of acquired user signals comprising;

generating a soft estimate of the complex transmit signal associated with the acquired user signal; and

refining or initially estimating at least one of a time offset estimate, a frequency offset estimate, a rate of change of frequency offset estimate, a complex gain estimate, an interference plus noise power estimate, or a signal to interference plus noise ratio (SINR) estimate using the soft estimate of the complex transmit signal, a previous soft estimate of the signal contribution from the user, and the residual signal to obtain an updated set of channel parameters for the acquired user signal and an updated soft estimate of the signal contribution in the residual signal from the acquired user signal;

d) updating the residual signal by subtracting the updated soft estimate of the signal contribution for each user signal obtained in step c) from the residual signal; and

e) performing another iteration of steps b) to d) until a termination condition is met wherein the updated residual signal in previous step d) is the residual signal in the next iteration of steps b) to d), and

wherein steps c) and d) may be performed serially, or step c) may be performed in parallel for each user signal in the set of acquired user signals, such that the residual signal for the next iteration of steps b) to d) comprises the residual signal from which the updated soft estimate of the signal contribution for each user signal has been subtracted, and during each iteration, step c) may be repeated for a user signal in the set of acquired users or step c) may be skipped for one or more user signals in the set of acquired user signals in which case the previous soft estimate of the signal contribution from the acquired user signal is used as the updated soft estimate of the signal contribution from the acquired user signal,

wherein the residual signal in step b) is the received signal from step a) if step d) has not yet been performed.

2. The method as claimed in claim 1, wherein in step b) searching for the predefined signal comprises using a coarse frequency estimator to obtain a coarse frequency estimate, and estimating the set of channel parameters for the acquired new user signal comprises performing fine frequency estimation to obtain an estimate of the frequency offset for the new user signal by adding the coarse frequency estimate and the fine frequency estimate.

3. The method as claimed in claim 2, wherein the predefined signal to aid acquisition is a discrete tone, and the coarse frequency estimator maximises a periodogram over a frequency range wherein the frequency range is restricted to allow for acquisition in specific frequency intervals or avoid acquisition in specific frequency intervals, and the coarse frequency offset is obtained from the frequency corresponding to the maximum absolute value of the periodogram by subtracting the frequency of the discrete tone.

4. The method as claimed in claim 1, wherein in step b) estimation of the set of channel parameters for the acquired new user signal takes into account that channel parameter values are correlated as a result of propagation mechanisms of a channel.

5. The method as claimed in claim 1, wherein prior to performing step a), a packet transmitted by a first user is received and decoded, and in step b) attempting to acquire one or more user signals comprises using information obtained from the previously received and decoded packet transmitted by the first user to acquire the first user as a new user signal in the residual signal.

6. The method as claimed in claim 1, wherein prior to performing step a), a packet transmitted by a first user is received and decoded, and in step b) attempting to acquire one or more user signals comprises using a knowledge of the system dynamics for the first user determined using the previously received and decoded packet transmitted by a first user to acquire the first user as a new user signal in the residual signal.

7. The method as claimed in claim 1, wherein in step b) attempting to acquire the one or more user signals in the residual signal comprises:

receiving an oversampled residual signal;

performing an initial compensation for a Doppler rate in the received oversampled residual signal;

estimating an initial coarse frequency offset using the predefined signal and removing an estimated coarse frequency offset from the oversampled residual signal;

estimating a time offset with respect to a nominal slot boundary;

filtering the oversampled residual signal using a pulse matched filter using the estimated time offset to generate a symbol rate signal;

estimating a complex signal gain, a fine frequency offset and frequency rate of change, wherein a final estimate of the frequency offset is obtained as the sum of the initial coarse frequency offset and the fine frequency estimate; and

removing or cancelling the predefined signal.

8. The method as claimed in claim 1, wherein for each new user signal acquired in step b), a new instance of a single user decoder module is spawned, and in step c) decoding of the residual signal for the acquired user signal is performed by the instantiated single user decoder module for the acquired user signal and the single user decoder module generates an updated soft estimate and the channel parameters of a current user signal and a hard decision of the data and comprises the following steps:

receiving an oversampled residual signal;

adding a previous soft estimate of the signal contribution from the current user signal to the oversampled residual signal;

removing an estimated effect of a communications channel on the current user signal by applying a previous estimate of the channel parameters for the current user signal to the received oversampled residual signal;

filtering the oversampled residual signal using a pulse matched filter to generate a symbol rate signal;

cancelling or removing the predefined signal;

estimating at least one of a signal amplitude, noise plus interference power, and signal to interference plus noise ratio (SINR);

soft decoding the symbol rate signal to generate a decoder quality metric, a data hard decision and a data soft decision;

soft remodulating data soft decisions and generating a soft symbol estimate for the user signal, and soft estimates of complex transmit signal;

adding the predefined signal to the complex transmit signal;

refining at least one of a time offset estimate, a frequency offset estimate, a rate of change of frequency offset estimate, a complex gain estimate, an interference plus noise power estimate, and a signal to interference plus noise ratio (SINR) estimate using the soft estimates of the complex transmit signal, the previous soft estimate of the signal contribution from the user signal, and the residual signal to obtain an updated set of channel parameters for the acquired user signal; and

applying the set of updated channel parameters to the soft estimate of the complex transmit signal to generate an updated soft estimate of the signal contribution from the user signal.

9. The method as claimed in claim 8, wherein the SINR is estimated by estimating a signal power from a most recent soft symbol estimate and an interference-plus-noise power is estimated from a symbol rate signal at the output of the pulse matched filter.

10. The method as claimed in claim 8, wherein the refining step comprises defining a search grid over one or more of the channel parameters and obtaining a frequency estimate at each grid point, and a best result over all grid points is used as a starting point for a numerical optimisation method to obtain an optimised channel parameter estimates for estimating the complex gain of the channel.

11. The method as claimed in claim 8, wherein the single user decoder module also generates a decoder quality metric, and in step c), if the decoder quality metric obtained in a previous iteration exceeds a predetermined threshold, then processing of the oversampled residual signal is skipped and the previous soft estimate of the signal contribution is output as the updated soft estimate of the signal contribution.

12. The method as claimed in claim 1, wherein decoding the residual signal, for each user signal in the set of acquired user signals, also generates a decoder quality metric for each user signal, and step b) further includes deciding whether to acquire a new user signal based on the decoder quality metrics of all previously acquired user signals.

13. The method as claimed in claim 1, wherein in step e) the termination condition is met if a confidence interval for an estimated channel parameter obtained in step b) exceeds a threshold range.

14. The method as claimed in claim 1, wherein the multiuser receiver is implemented as a distributed receiver, wherein processing is segmented into two stages, where the outputs of the first stage feed a second stage via a network.

15. The method as claimed in claim 14, wherein the distributed receiver comprises multiple receive antennas and first stage processors, and the receiver second stage operates for one or more iterations on inputs coming from multiple first stages in series or in parallel.

16. The method as claimed in claim 1, wherein the multiuser receiver is located on a satellite and the multiple transmission are over a satellite channel with significant frequency shifts, frequency rate of change or time delays compared to a terrestrial channel.

17. A multiuser decoder for use in a multiuser receiver in a slot based wireless communication system, the multiuser decoder configured to decode multiple transmissions in a residual signal received in a slot using iterative soft interference cancellation, the multiuser decoder comprising:

an initial acquisition module configured to acquire one or more user signals in a residual signal by searching for a predefined signal inserted in a complex transmit signal transmitted by a user to aid acquisition, and to estimate a set of channel parameters for each acquired user signal, wherein the set of channel parameters comprises at least one of a complex gain, a time offset, a frequency offset, a rate of change of frequency offset, or a noise variance; and

one or more single user decoder modules wherein each single user decoder is configured to generate a soft estimate of the signal contribution from an acquired user signal in the residual signal and to update the set of channel parameters for the acquired user signal, and comprising;

a signal to interference plus noise ratio (SINR) estimator to estimate at least one of a signal amplitude, noise plus interference power, and SINR;

a soft decoder;

a soft remodulator configured to generate soft estimates of a complex transmit signal associated with the acquired user signal;

a channel parameter updater block configured to refine or initially estimate at least one of a time offset estimate, a frequency offset estimate, a rate of change of frequency offset estimate, and a complex gain estimate, an interference plus noise power estimate, or a signal to interference plus noise ratio (SINR) estimate using the soft estimates of the complex transmit signal, a previous soft estimate of the signal contribution from the user, and the residual signal to obtain the updated set of channel parameters;

a channel application module configured to apply the updated channel parameters to the soft estimates of the complex transmit signal to generate a soft estimate of the signal contribution in the residual signal from the acquired user signal; and

wherein the multiuser decoder is configured to iteratively generate an updated residual signal until a termination condition is met wherein each iteration comprises acquiring the one or more user signals in the residual signal using the initial acquisition module, and for each acquired user signal, spawning an instance of a single user decoder module, and then for each acquired user signal, the respective single user decoder generates the soft estimate of the signal contribution from the acquired user signal, and then generating the updated residual signal by subtracting the soft estimate of the signal contribution for each acquired user signal obtained from each single user decoder from the residual signal, and in the first iteration the residual signal provided to the initial acquisition module is a received signal and in subsequent iterations the residual signal provided to the initial acquisition module is the updated residual signal,

wherein the one or more single user decoder modules may be configured serially or in parallel, and during each iteration, a single user decoder module for a user in the set of acquired users may be repeatedly activated or may be skipped in which case the previous soft estimate of the signal contribution from the acquired user signal is used as the updated soft estimate of the signal contribution from the acquired user signal.

18. An initial acquisition module for use in a multiuser decoder configured to acquire one or more user signals in a residual signal, the initial acquisition module comprising:

an input for receiving an oversampled residual signal;

a compensator performing an initial compensation for a Doppler rate in the received oversampled residual signal;

a coarse frequency offset estimator for estimating an initial coarse frequency offset by searching for a predefined signal inserted in a complex transmitted signal transmitted by a user to aid acquisition and the coarse frequency offset estimator is configured to remove the estimated initial coarse frequency offset from the residual signal;

a time offset estimator for estimating a time offset with respect to a nominal slot boundary;

a pulse matched filter for filtering the oversampled residual signal using the estimated time offset from the time offset estimator to generate a symbol rate signal;

an estimator module for estimating a complex signal gain, a fine frequency offset and frequency rate of change from the symbol rate signal, wherein a final estimate of the frequency offset is obtained as a sum of the initial coarse frequency offset and a fine frequency estimate; and

an acquisition signal removal module for either removing or cancelling the predefined signal inserted into the transmitted signal by the user to aid acquisition.

19. A single user decoder module configured to provide updated versions of a soft estimate of the signal contribution from a user signal in a signal and to provide updated channel parameters for the user signal and hard decisions of data, the single user decoder comprising:

an input for receiving an oversampled residual signal;

an adder for adding a previous soft estimate of the signal contribution from the user signal;

a channel removal module for removing an estimated effect of a communications channel on the user signal by using a previous estimate of the channel parameters;

a pulse matched filter for filtering the oversampled residual signal to generate a symbol rate signal;

an acquisition signal removal module for either removing or cancelling a predefined signal inserted into a transmitted signal by a user to aid acquisition from the oversampled residual signal;

a signal to interference plus noise ratio (SINR) estimator to estimate at least one of a signal amplitude, noise plus interference power, and SINR;

a decoder for soft decoding the oversampled residual signal to generate a decoder quality metric, a data hard decision and a data soft decision;

a soft remodulator for soft remodulating the data soft decision and generating a soft symbol estimate for the user signal, and soft estimates of a complex transmit signal associated with the received oversampled residual signal;

an acquisition signal adder module for adding the predefined signal to aid acquisition to the complex transmit signal;

a channel parameter updater block configured to refine at least one of a time offset estimate, a frequency offset estimate, a rate of change of frequency offset estimate, and a complex gain estimate, an interference plus noise power estimate, and a signal to interference plus noise ratio (SINR) estimate using the soft estimates of the complex transmit signal, a previous soft estimate of the signal contribution from the user, and the residual signal to obtain updated channel parameters; and

a channel application module configured to apply the updated channel parameters to the soft estimates of the complex transmit signal to generate a soft estimate of the signal contribution from the user signal.