Qr-rls adaptive digital filter with 18-bit pipeline architecture

A method for communicating information enabling a receiver to reliably determine if a packet has been received, and enabling the receiver to reliably identify a modulation scheme used to encode a received packet, the method comprising:

• assigning to each packet a group of X modulation bits located near the beginning of the packet;
• assigning to each of a plurality of available modulation schemes a modulation value, each modulation value being in one of two ranges, the two ranges being −X to −X/Y and +X/Y to +X, where Y is a number selected so as to create a desired gap between the two ranges;
• encoding by a transmitter into the modulation bits the modulation value for a selected available modulation scheme using a standard modulation scheme, a remainder of the packet being encoded by the transmitter using the selected available modulation scheme, the modulation value being represented as a sum of modulation bits such that each of the summed modulation bits has equal significance;
• detecting of the packet by the receiver, each of the X modulation bits being detected by b weight bits, so that the two ranges are represented by two ranges of detected modulation values, the two ranges of detected modulation values being the ranges −2b-1Nr to −(2b-1Nr)/Y and (2b-1Nr)/Y to 2b-1 Nr, where r is the value of a fully confident bit estimate, the modulation values being surrounded by non-overlapping regions of values which are bounded by threshold values;
• summing of the detected modulation bits so as to determine a decision statistic;
• comparing the decision statistic with the threshold values;
• and
• if the decision statistic falls within a region corresponding to an available modulation scheme, determining that the modulation scheme of the packet is the corresponding modulation scheme, otherwise, determining that a valid packet has not been received.

A method for removing an impulse signal from an input stream of a digital filter, the method comprising:

• obtaining a baseline sample set from the input stream and calculating a baseline mean square estimate (B-MSE) thereof, the baseline sample set including sufficient samples to ensure that the B-MSE will not be significantly affected if an impulse is contained therein;
• dividing the baseline sample set into a plurality of candidate impulse sets;
• for each of the candidate impulse sets, calculating an impulse mean square estimate (I-MSE) thereof, and declaring that the candidate impulse set includes an impulse if the ratio I-MSE/B-MSE exceeds an impulse threshold value X;
• and
• if the candidate impulse set includes an impulse, removing the impulse from the data stream.

A genetic method for identifying groups of asynchronous spreading codes having minimal mutual correlations, the method comprising:

• randomly selecting a population of candidate spreading codes;
• grouping the candidate spreading codes into competition groups;
• within each competition group, ranking the spreading codes according to their fitness relative to the other members of the competition group;
• for each competition group, eliminating at least the candidate spreading code having the lowest fitness;
• replacing the eliminated candidate spreading codes with randomly selected replacement spreading codes;
• and
• repeating the steps of grouping, ranking, eliminating, and replacing until a termination criterion has been met.

A method for determining a sampling window appropriate for an input signal supplied to a QR-RLS filter, the input signal including a plurality of streams of overlapping symbols arising from a plurality of users, the symbols having relative timing delays TD(i), the timing delays TD(i) being at least approximately known, the method comprising:

• sorting the input streams according to their timing delays from lowest to highest and determining the largest timing delay, also referred to herein as the maximum timing delay TDmax;
• increasing the timing delays of all of the input streams in units of one symbol so that all of the timing delays are greater than TDmax;
• re-sorting the input streams according to their adjusted timing delays from lowest to highest;
• assigning to each user an integer index i;
• setting a minimum delay MD variable and a time offset TO to initial values of one symbol length each;
• iterating for each user i the steps of:
• subtracting TD(i) from TDmax to determine an offset O(i);
• increasing TD(i) by one symbol length;
• and
• if O(i) is less than MD, resetting MD equal to O(i) and setting the time offset TO to TD(i);
• and
• defining the beginning of the sampling window as the value of TO resulting from the step of iterating.