Methods and systems for synchronizing wireless transmission of data packets (05-Apr-2011)

CROSS REFERENCE TO RELATED APPLICATION

The present application claims priority under 35 U.S.C. §119 to a corresponding patent application filed in India and having application number 269/CHE/2010, filed on Feb. 3, 2010, the entire contents of which are herein incorporated by reference.

BACKGROUND

In a wireless communication system, a data packet format usually includes control information and user data. The control information includes data such as source and destination addresses, error detection codes like checksums, and sequencing information. Control information may be found in packet headers and trailers, and user data is included in between. The control information further includes a fixed-pattern preamble. The preamble serves several purposes, namely, to identify a packet type, and to provide a signal for Direct Current (DC) offset estimation, frequency offset estimation and channel estimation, for example.

In addition, another example purpose of the preamble is to allow time for a receiver to achieve lock of a receiver digital phase lock loop that is used to synchronise a receive data clock to a transmit data clock. At a point when a first bit of the preamble is received, a receiver may be in an arbitrary state (i.e., have an arbitrary phase for a local clock of the receiver). During the course of the preamble, the receiver learns a correct phase, but in so doing, the receiver may miss (or gain) a number of bits. A preamble thus usually includes a predetermined pattern to mark the last two bits of the preamble. When the pattern is received, the receiver begins collecting bits into bytes for data processing. The receiver may also confirm a polarity of a transition representing a logic high data bit to the receiver (as a check in case the bit has been inverted), for example.

Different communications protocols use different conventions for distinguishing between control information and user data. In Binary Synchronous Transmission, for example, the data packet is formatted in 8-bit bytes, and special characters are used to delimit different elements. Other protocols, like Ethernet, establish a beginning of a header and data elements by a location relative to the start of the data packet. Some other protocols format information at a bit level instead of a byte level.

Many over-the-air modulation techniques use similar basic protocols, and each technique usually includes use of preambles for use in transmission of data as well as management and control of wireless links. Transmitters and receivers may be programmed and designed to receive wireless signals, and to decode the signals based on expected preamble content. However, as new technology is developed, new modulation techniques may not be compatible with previous transmitter and receiver designs. Thus, for new and legacy systems to co-exist, new modulation techniques may need to have the ability to generate legacy data packets for the legacy systems and high throughput packets for new systems. To do so, modified preambles may be employed to enable both legacy equipment and new equipment to detect information in a received data packet.

SUMMARY

In an example aspect, a method for synchronizing wireless transmission of data packets is provided. The method includes receiving a plurality of signals, and each signal has a data packet frame format including a preamble and data. The preamble comprises a pattern of symbol sequences repeated a number of times. The method also includes correlating patterns of symbol sequences within the plurality of signals to produce correlation values. The method also includes calculating a first metric of the plurality of signals. The first metric is a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals. The method also includes calculating a second metric that defines an average of a power of noise in the plurality of signals. The method further includes determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

In another example aspect, a computer readable medium having stored therein instructions executable by a computing device to cause the computing device to perform functions is provided. The functions include receiving a plurality of signals and each signal has a data packet frame format including a preamble and data. The preamble comprises a pattern of symbol sequences repeated a number of times. The functions further include correlating patterns of symbol sequences within the plurality of signals to produce correlation values. The functions also include calculating a first metric of the plurality of signals that is a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals. The functions also include calculating a second metric that defines an average of a power of noise in the plurality of signals, and determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

In another example aspect, a system is provided that comprises a processor, a data storage medium, and machine language instructions stored on the data storage medium and executable by the processor to perform functions including receiving a plurality of signals and each signal has a data packet frame format including a preamble and data. The preamble comprises a pattern of symbol sequences repeated a number of times. The functions also include correlating patterns of symbol sequences within the plurality of signals to produce correlation values, and calculating a first metric of the plurality of signals that is a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals. The functions also include calculating a second metric that defines an average of a power of noise in the plurality of signals, and determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

In another example aspect, each signal of a plurality of signals includes multiple versions of a transmitted signal, and each version is a cyclically shifted version of the transmitted signal. The example methods and example systems may thus further include functions of receiving the plurality of signals at a plurality of receive antennas, and at each receive antenna, determining a cross correlation between the multiple versions of the transmitted signal to produce a number of peaks equal to a number of transmit antennas where each peak corresponds to a position to where the transmitted signal was shifted. The functions further include at each receive antenna, combining cross correlations of the multiple versions of the transmitted signal by shifting correlated outputs so that peaks substantially match to produce combined cross-correlations, adding combined cross-correlations from each antenna to produce a third metric, and identifying a position of a maximum peak within the third metric as a fine timing offset amount that defines the approximate division of the preamble and the data in the data packet frame format of each signal. In one example, the approximate division may be an optimal division of the preamble and the data in the data packet.

The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of an example multiuser MISO system.

FIG. 2 is a block diagram of an example MIMO-OFDM transmitter.

FIG. 3 illustrates an example frame format of a data packet for a non-HT frame format.

FIG. 4 illustrates an example frame format of a data packet for a HT mixed frame format.

FIG. 5 illustrates an example frame format of a data packet for a HT only frame format.

FIG. 6 is a flowchart that depicts example steps of a method for determining a metric used in a coarse timing offset estimation.

FIG. 7 is a flowchart that depicts example steps of a method for determining another metric used in a coarse timing offset estimation.

FIG. 8 is an example plot of the metrics determined from FIGS. 7 and 8.

FIG. 9 is a block diagram illustrating an example coarse timing offset estimator.

FIG. 10 is a flowchart that depicts example steps of a method for determining a fine timing offset estimation of a beginning of data within a data packet.

FIG. 11 is a block diagram illustrating an example fine timing offset estimator.

FIGS. 12-15 illustrate example plots of simulations.

FIG. 16 is a block diagram illustrating an example computing device arranged for computing a coarse timing offset and/or a fine timing offset estimation of a beginning of data in a data packet.

DETAILED DESCRIPTION

In the following detailed description, reference is made to the accompanying drawings, which form a part hereof. In the drawings, similar symbols typically identify similar components, unless context dictates otherwise. The illustrative embodiments described in the detailed description, drawings, and claims are not meant to be limiting. Other embodiments may be utilized, and other changes may be made, without departing from the spirit or scope of the subject matter presented here. It will be readily understood that the aspects of the present disclosure, as generally described herein, and illustrated in the Figures, can be arranged, substituted, combined, and designed in a wide variety of different configurations, all of which are explicitly contemplated and are made part of this disclosure.

In example embodiments below, methods are described for synchronizing wireless transmission of data packets. The institution of electrical and electronics engineers (IEEE) 802.11 family of specifications describes over-the-air modulation techniques that use the same basic protocol, and each technique includes use of preambles for use in transmission of data as well as management and control of wireless links. In IEEE 802.11n wireless local area network (WLAN) systems, which employ multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM), there are two modes of transmission that are used (e.g., a mixed mode transmission and a green field mode of transmission) and each mode employs a different preamble, for example. Methods and systems described herein define techniques to detect preambles and data within a data packet.

Example methods described herein include receiving a plurality of signals, and each signal has a data packet frame format including a preamble and data. The preamble comprises a pattern of symbol sequences repeated a number of times. The methods further include correlating patterns of symbol sequences within the plurality of signals to produce correlation values. A first metric of the plurality of signals can be calculated that is a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals. In addition, a second metric can be calculated that defines an average of a power of noise in the plurality of signals. The methods further include determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric. The methods may further include determining a cross-correlation between versions of a transmitted signal to produce a number of peaks equal to a number of transmit antennas, generating a third metric based on combinations of cross-correlations, and identifying a position of a maximum peak within the third metric as a fine timing offset amount.

Example embodiments described below may be arranged, substituted, combined, and designed in a wide variety of different configurations, all of which are explicitly contemplated and are made part of this disclosure.

In wireless communications, a single antenna may be used at a source to wirelessly transmit a signal to a single antenna at a destination. In some cases, this may give rise to problems due to multipath effects. For example, when a wireless signal (an electromagnetic field) is met with obstructions such as hills, canyons, buildings, and utility wires, wavefronts are scattered, and thus may take many paths to reach the destination. Late arrival of scattered portions of the signal causes problems such as fading, cut-out, and intermittent reception.

Multiple input multiple output (MIMO) systems can be used in which multiple antennas are used at both the source (transmitter) and the destination (receiver). The antennas at each end of the communication are combined to minimize errors and optimize data speed. MIMO is one of several forms of smart antenna technology. Other forms include multiple input, single output (MISO) and single input, multiple output (SIMO). Using smart antenna technology (e.g., multiple antennas at both the source and the destination) can eliminate the signal problems caused by multipath wave propagation, and can even take advantage of this effect.

FIG. 1 is a block diagram of an example multiuser MISO system 100 including a base station (BS) 102 communicating with N_u users on a downlink communication channel 104. The BS 102 includes transmit antennas 1, 2, . . . , N_t, and each downlink user is equipped with one receive antenna, such as antennas 106, 108 and 110.

As one example of a MIMO system, the IEEE 802.11n standard, which employs multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM), uses different operating modes configured based on a type of device in the network. The different operating modes include non-high throughput (non-HT) mode, high throughput (HT) mixed mode, and HT green field mode.

For example, in the mixed mode of transmission, both 802.11n MIMO-OFDM systems and legacy 802.11 systems co-exist, and the 802.11n MIMO system has the capability to generate legacy data packets for the legacy systems and high throughput packets for MIMO-OFDM systems. Thus, in the mixed mode of transmission, a long header preamble is employed to enable both legacy equipment and 802.11n equipment to detect information in the header. In the green field mode of transmission, only 802.11n MIMO-OFDM systems are used. The green field mode is intended for high throughput transmission only. No transmissions are intended to legacy equipment or mixed mode systems from the green field system. In the green field mode of transmission, a shortened header can thus be used and is only applicable to a selected set of equipment for synchronization purposes. Data packets transmitted in the green field mode will have only MIMO specific preambles, and thus, no legacy format preambles are present. Because of the different types of preambles used in 802.11n networks for the different modes of transmission, different methods are used for synchronization purposes in the 802.11n networks that support both preambles. The MIMO-OFDM receivers should be able to decode the green field mode packets as well as legacy format packets.

An example MIMO-OFDM transmitter 200 used in the HT mixed mode of the 802.11n standard is shown in FIG. 2. The transmitter 200 may be a base station, such as shown in FIG. 1, and each block may represent a module, a segment, or a portion of program code, which includes one or more instructions executable by a processor for implementing specific logical functions or steps in the process. In addition, each block may represent circuitry that is wired to perform the specific logical functions.

Incoming data bits to be transmitted by the transmitter 200 are randomized with a unique scramble code using a scrambler 202 to avoid occurrence of long zeros and ones. The unique scramble code is provided to a receiver, so that the receiver may multiply a received signal by the same scramble code that the transmitter used in order to descramble bits of the signal, for example.

A forward error correction encoder (FEC) block 204 encodes the scrambled data to enable channel error correction capabilities. The FEC block 204 includes a binary convolutional encoder followed by a puncturing block, for example. A basic coding rate of ½ is achieved using the FEC block 204 and other coding rates, such as ¾, ⅔, and ⅚ are achieved with the help of a puncturing pattern defined in the 802.11n standard.

All encoded data bits are interleaved by a block interleaver 206 with a block size corresponding to a number of bits in a single OFDM symbol, N_CBPS (number of coded bits per symbol). The interleaver 206 is defined by a two-step permutation. A first permutation ensures that adjacent coded bits are mapped onto nonadjacent subcarriers. A second permutation ensures that adjacent coded bits are mapped alternately onto less and more significant bits of the constellation, and thereby, long runs of low reliability least significant bits (LSB) are avoided. Then, the interleaved bits are mapped to complex symbols using a quadrature amplification modulation (QAM) constellation mapper 208. Out of 64 subcarriers in 20 MHz operation, for example, data symbols are mapped onto subcarriers −26 to −1 and +1 to 26. Remaining subcarriers are loaded with guard subcarriers and pilot symbols, for example.

Then an N-point inverse fast Fourier transform (IFFT) using an IFFT block 210 is taken and distributed across N_t transmit (TX) antennas in a circular manner. In each transmit antenna, a cyclic shift delay (CSD) element 212a-212t is introduced across the signals. This helps avoid a beamforming effect at a receiver during preamble transmission. A cyclic prefix of length N_cp is appended in front of a cyclically shifted IFFT output to combat intersymbol interference caused due to a multipath channel. It is assumed that a total TX power is divided equally among the N_t transmit antennas. The signal is then upconverted to radio frequency using a cyclic prefix (CP) and radio frequency (RF) block 214a-214t and transmitted through the spatially correlated quasi-static multipath channel using an antenna 216a-216t. Thus, the antennas 216a-216t each transmit the same data packet although shifted in time by the CSD elements 212a-212t.

In the non-HT operating mode of 802.11n, a format of a transmitter will be similar to the transmitter 200 shown in FIG. 2, except that the transmitter will include a single transmit antenna chain. A cyclical delay shift (CDS) applied to a transmit chain in a non-HT transmitter is 0 μs, for example, since there is only a single transmit antenna chain.

In an HT green field operating mode of 802.11n, a format of a transmitter is different from the mixed mode transmitter 200 and is defined in such a way that improved performance is achieved in MIMO only configurations. For example, incoming data bits are scrambled bits and then parsed into 1 or 2 sequences for encoding. The encoded sequences are passed into a single stream parser where the data is parsed into N_t data streams, and in each stream interleaving, constellation mapping, and CSD are applied separately. All the data streams are then passed through a spatial mapper where mapping of data symbols into data subcarriers of N_t inverse discrete Fourier transform (IDFT) blocks is performed. After the IDFT operation and CP appending, the signal is upconverted and transmitted in the RF.

The CSD introduced into the data packet for each transmit chain of the non-HT and HT portion of the preamble is given in Tables 1 and 2 below in terms of samples and time in nanoseconds (ns). For example, a transmitter may have multiple antennas, and each antenna may transmit the same data packet although shifted in time by the cyclic shift shown below. A receiver will receive all signals, and starting positions of the signals will be different due to the cyclic shift, for example.

TABLE 1
Cyclic shift for the non-HT portion of the packet

Number of	Cyclic shift	Cyclic shift	Cyclic shift	Cyclic shift
Transmit	for TX	for TX	for TX	for TX
Chain	chain 1 (ns)	chain 2 (ns)	chain 3 (ns)	chain 4 (ns)

1	0
2	0	−200
3	0	−100	−200
4	0	−50	−100	−150

Example cyclic shifts for non-HT portions of data packets are shown in Table 1. For example, as shown in Table 1, a cyclic shift for a transmit antenna chain with only one antenna is zero because there can be no shifted signal sent by a second antenna. For a transmit chain of two antennas, a second antenna shifts the signal by 200 ns (e.g., delays transmitting the signal for 200 ns after transmission of the signal by the first antenna). For a transmit antenna chain of three antennas, the second antenna delays transmission of the signal by 100 ns after transmission by the first antenna, and the third antenna delays transmission by 200 ns after transmission by the second antenna, for example. For a transmit antenna chain of four antennas, the second antenna delays transmission the signal by 50 ns after transmission by the first antenna, the third antenna delays transmission by 100 ns after transmission by the second antenna, and the fourth antenna delays transmission by 150 ns after transmission by the third antenna, for example.

TABLE 2
Cyclic shift for the HT portion of the packet

Number of	Cyclic shift	Cyclic shift	Cyclic shift	Cyclic shift
Transmit	for TX	for TX	for TX	for TX
Chain	chain 1 (ns)	chain 2 (ns)	chain 3 (ns)	chain 4 (ns)
1	0
2	0	−400
3	0	−400	−200
4	0	−400	−200	−600

Example cyclic shifts for HT portions of data packets are shown in Table 2. For example, as shown in Table 2, a cyclic shift for a transmit antenna chain with only one antenna is zero because there can be no shifted signal sent by a second antenna. For a transmit chain of two antennas, a second antenna shifts the signal by 400 ns (e.g., delays transmitting the signal for 400 ns after transmission of the signal by the first antenna). For a transmit antenna chain of three antennas, the second antenna delays transmission of the signal by 400 ns after transmission by the first antenna, and the third antenna delays transmission by 200 ns after transmission by the second antenna, for example. For a transmit antenna chain of four antennas, the second antenna delays transmission the signal by 400 ns after transmission by the first antenna, the third antenna delays transmission by 200 ns after transmission by the second antenna, and the fourth antenna delays transmission by 600 ns after transmission by the third antenna, for example.

Table 3 below provides cyclic shift duration values depicted according to a corresponding cyclic shift in samples. The negative values in the tables indicate a shift to the left, for example.

	TABLE 3
	Cyclic shift duration	Cyclic shift in
	in nanoseconds	number of samples (−k)

	0	0
	−50	−1
	−100	−2
	−150	−3
	−200	−4

Different operating modes of wireless protocols may have specific frame formats, and for backward compatibility reasons, new technology formats (such as the HT MIMO-OFDM data format of IEEE 802.11n) may use specific fields of non-HT preambles, for example. FIGS. 3-5 illustrate example frame formats of data packets for different operating modes of an example system.

FIG. 3 illustrates an example frame format of a data packet for a non-HT frame format. The frame format of the data packet shown in FIG. 3 may be used in legacy networks where only 802.11a/g enabled devices are present, for example. The data packet includes a number of short symbols (SS) each of a specific duration, and the short symbols are referred to as the short training field (STF) 302. For example, the data packet may include 10 identical short symbols each of a duration of about 0.8 μs in the STF 302. Initial receiver tasks, such as packet arrival detection, automatic gain control (AGC), coarse time acquisition, and coarse frequency acquisition may be performed using a correlation property of the STF, for example.

The data packet also includes an extended cyclic prefix (CP), which is followed by long symbols (LS). The CP and the long symbols are referred to as a long training field (LTF) 304. The CP may be of a duration of about 1.6 μS, and the LTF 304 may include two identical long symbols (LS) each of a duration of about 3.2 for example. The LTF is used for channel estimation, fine timing, and fine frequency acquisition, for example.

Following the LTF 304, a signal field (SIG) carries rate and length information of the data packet. The SIG field may be of a duration of about 4 μs, for example. After the SIG field, data begins.

FIG. 4 illustrates an example frame format of a data packet for a HT mixed frame format. The frame format of the data packet shown in FIG. 4 may be used in a mixed mode network where HT mobile stations and legacy mobile stations co-exist, for example. For example, in a network including multiple antenna enabled systems, the antennas may be capable of transmitting and receiving legacy frames and MIMO-OFDM frames, for example.

To provide backward compatibility for legacy systems, the data packet in FIG. 4 includes all three fields of the non-HT preamble as shown in FIG. 3 including an STF, an LTF, and an SIG field. For example, the data packet in FIG. 4 includes a legacy format preamble 402 including the STF in the form of legacy short training fields (L-STF), the LTF in the form of legacy long training fields (L-LTF), and the SIG field in the form of a legacy signal field (L-SIG). Providing the legacy format preamble 402 enables legacy mobile stations to decode the SIG field of an HT frame format, and allows MIMO-OFDM transmissions to progress without collisions, for example.

The data packet in FIG. 4 further includes a high throughput preamble 404 including a high throughput signal field (HT-SIG), a high throughput short training field (HT-STF), and N_t high throughput long training fields (HT-LTF). Following the high throughput preamble 404 is data to be decoded. The N_t HT-LTF fields are used for estimating a channel between multiple transmit and multiple receive antennas, for example. When transmission of the data packet is intended for a MIMO-OFDM receiver, then based on an antenna configuration, the data packet frame format shown in FIG. 4 is transmitted after applying the cyclic shift as described above in Tables 1 and 2 for non-HT and HT portions, respectively.

FIG. 5 illustrates an example frame format of a data packet for a HT only frame format. The frame format of the data packet shown in FIG. 5 may be used in a high throughput network only, for example, such as in the greenfield network mode of IEEE 802.11n where only HT MIMO-OFDM mobile stations operate. The data packet frame format of FIG. 5 includes a high throughput preamble 502 including training fields for MIMO-OFDM systems, for example. The training fields include a high throughput short training field (HT-STF), a high throughput long training field (HT-LTF), a high throughput (HT-SIG) field, and high throughput long training fields (HT-LTFs) in sequence. Following the high throughput preamble 502 is data to be decoded.

The HT-STF field may be the same as the L-STF in the legacy format preamble 402 of the mixed mode data packet frame format, and can be used for timing acquisition, AGC and frequency acquisition, for example. To demodulate content of the HT-SIG, channel estimates can be obtained from the HT-LTF field, for example. For transmit antennas, fields in the data packet frame format are cyclically shifted by using delays described in Table 2, and then transmitted, for example.

In example embodiments, a transmitter transits information without informing a receiver when information or data will begin in the data packet, and the receiver may use patterns sent by transmitter to determine a starting point of the data. In addition, in a wireless network many transmitters and receivers may use and interpret modified preambles, while legacy receivers may not be expecting modified preambles, but might receive signals including such preambles. Thus, due to different transmission modes, a method for determining a starting point of the data or synchronizing data transmission is provided.

A received signal at an r^th receive antenna is given as:

y r ⁡ ( n ) = ∑ t = 0 N t - 1 ⁢ ⁢ ∑ p = 0 P - 1 ⁢ ⁢ x t ⁡ ( n ) ⁢ h rt ⁡ ( n - p ) ⁢ ⅇ j2πɛ ⁢ ⁢ n + v r ⁡ ( n ) Equation ⁢ ⁢ ( 1 )
where h_rt(n) is an impulse response of a channel between the r^th receive antenna and the t^th transmit antenna, x_t(n) is a transmit signal at an instant n from the t^th transmit antenna, and ν_r(n) is additive white Gaussian noise (AWGN) at the r^th receive antenna with a variance of σ_ν². In addition, p is a length of a multipath channel and remains static across n. The received signal is affected by normalized frequency offset ε caused due to mismatch in oscillators between a transmitter and a receiver.

The received signal includes a preamble and data (e.g., as shown in the data packet frame formats of FIGS. 3-5). In example embodiments, a beginning of the data in the data packet is determined. Example methods described below include determining a rough estimation of a beginning of the data, and determining an exact estimation of the beginning of the data in the data packet.

The data packet includes a preamble comprising a structure or training sequence. The training sequences (STF and LTF) are repetitive patterns of the same sequence transmitted at the beginning of the start of transmission. The repetitive patterns may be defined according to a standard protocol, for example, so that a receiver knows what pattern to search for within a received signal. For example, for the non-HT frame format shown in FIG. 3, the LTF (legacy field) can be detected by old and new mobile stations, and for the HT mixed frame format shown in FIG. 4, a legacy format preamble 402 is included as well so that legacy mobile stations may also detect such a preamble. Using example embodiments provided below, a legacy mobile station may be able to detect and use preambles of formats shown in FIGS. 3-4, for example.

Coarse Timing Offset Estimation Method

After detecting a data packet arrival, automatic gain control (AGC) is triggered and performed (e.g., an average output signal level is fed back to adjust a gain to an appropriate level for a range of input signal levels). An example coarse time offset (CTO) estimator then finds a rough starting position of any of the short symbols (e.g., as shown in FIG. 3 in the STF 302) in the received data packet. An end of the STF can be determined by using the autocorrelation property of the received signal. For example, the STF comprises a pattern (SS) of 16 samples repeated 10 times according to the IEEE 802.11n standard. Each pattern of 16 samples “N_s” in the STF is highly correlated with another pattern, and that can be exploited at the receiver. Followed by the STF, an LTF is transmitted. These two training sequences that are transmitted at the start of the communication have been designed to maintain synchronization of the receiver. Other wireless protocols may also transmit other training patterns as well.

To determine a start of one of the patterns in the STF, autocorrelation of the sequence with a lag equal to the periodicity (e.g., N_s in this case) in the STF is used. The receiver will determine correlation values between patterns that are separated by N_s samples, and the correlation values will be high during the STF reception period when compared to the LTF and/or data OFDM symbols reception period. At the end of the STF reception period, the correlation values begin to decrease as the periodic pattern does not hold for subsequent signals. This enables autocorrelation to be used to determine a rough end of the STF. Note that more operations can be performed to determine an exact end of the STF, however, in example embodiments, a coarse (rough) timing offset estimation is sufficient.

The cyclic shift applied to the signal does not affect the autocorrelation because each pattern in a received signal will be the same, only shifted in time. Thus, the autocorrelation principles will apply, for example.

Using the autocorrelation method, only a rough estimate of the start of the STF may be determined because a received signal may be affected due to channel impairments, receiver frontend noise, etc., that can add certain distortion effects and noise. Thus, an exact repetition of the pattern may not be seen at the receiver, however, highly correlated data can be identified and used as a reliable timing point (start or end of sequence), for example.

Let k be the instant at which the AGC is converged and samples are fed through a coarse time offset (CTO) estimator (described below in FIG. 9). A metric S(n) is calculated from the instant k at which the AGC is converged. The metric is given by:

S ⁡ ( n ) = 1 N r ⁢ ∑ r = 0 N r - 1 ⁢ ⁢  P r ⁡ ( n )   R r ⁡ ( n )  Equation ⁢ ⁢ ( 2 )
where

P r ⁡ ( n ) = ⁢ ∑ d = 0 N s - 1 ⁢ ⁢ y r ⁡ ( n + d ) ⁢ y r * ⁡ ( n + N s + d ) = ⁢ ∑ d = 0 N s - 1 ⁢ ⁢ ∑ t = 0 N t - 1 ⁢ ⁢ ∑ p = 0 P - 1 ⁢ ⁢  x t ⁡ ( n + d - p )  2 ⁢  h rt ⁡ ( p )  2 + α r ⁡ ( n ) Equation ⁢ ⁢ ( 3 )
and

R r ⁡ ( n ) = ⁢ 1 N s ⁢ ∑ d = 0 N s - 1 ⁢ ⁢ y r ⁡ ( n + N s + d ) ⁢ y r * ⁡ ( n + N s + d ) = ⁢ 1 N s ⁢ ∑ d = 0 N s - 1 ⁢ ⁢ ∑ t = 0 N t - 1 ⁢ ⁢ ∑ p = 0 P - 1 ⁢ ⁢  x t ⁡ ( n + d - p )  2 ⁢  h rt ⁡ ( p )  2 + ⁢ β r ⁡ ( n ) Equation ⁢ ⁢ ( 4 )
Equation (3) defines the autocorrelation of a received signal, and Equation (4) defines a power estimate of a received signal. While calculating the autocorrelation (P_r(n)) and the power estimate of the received signal (R_r(n)), an impact of channel and frequency offset is nullified because of a sliding window autocorrelation performed on the received signal with a window length of N_s samples. In Equation (3), α_r(n) is a value of the cross correlation between the signal and noise terms. Similarly, in Equation (4), β_r(n) is a sum of noise energy and value of cross correlation between the signal and noise terms.

Thus, using Equation (3), an autocorrelation of the signal can be determined and is used in the metric defined in Equation (2). The value of metric S(n) can take different values based on the index n. For example,
S(n)≦1∀n≦8N_s
S(n)<1∀n>8N_s+K with 0<K<N_s.
The parameter P_r(n) can be rewritten as:

P r ⁡ ( n ) = ∑ d = 0 N s - K - 1 ⁢ ⁢ ∑ t = 0 N t - 1 ⁢ ⁢ ∑ p = 0 P - 1 ⁢ ⁢  x t ⁡ ( n + d - p )  2 ⁢  h rt ⁡ ( p )  2 + α r ⁡ ( n ) Equation ⁢ ⁢ ( 5 )
In Equation (5), α_r(n) is a sum of the cross correlation of the signal and noise terms, and cross correlation between samples from the STF and the LTF. Since the fields STF and LTF are highly uncorrelated, the parameter α_r(n) decreases with n>8N_s, which reduces P_r(n). The metric S(n) will form an end of the plateau and could be noisy due to AWGN and multipath fading conditions. To have a smooth plateau, the metric is filtered through a weight filter and is given as:
S′(n)=γS′(n−1)+(1−γ)S(n)  Equation (6)
where γ is the weight factor given to previous value and (1−γ) is a weight applied to the current metric.

Equation (6) refines the metric by filtering the metric to remove noise. In addition, because of possible multiple channel distortion and noise, the metric might not be smooth. Thus, a smoothing function is also performed by taking a previous sample, and applying a weight with a factor (γ), and a factor (1−(γ)) as the weight given to a current sample, for example.

FIG. 6 is a flowchart that depicts example steps of a method for determining a metric used in a coarse timing offset estimation for a beginning of data within a data packet. It should be understood that the flowchart shows functionality and operation of one possible implementation of present embodiments. In this regard, each block may represent a module, a segment, or a portion of program code, which includes one or more instructions executable by a processor for implementing specific logical functions or steps in the process. The program code may be stored on any type of computer readable medium, for example, such as a storage device including a disk or hard drive. In addition, each block may represent circuitry that is wired to perform the specific logical functions in the process. Alternative implementations are included within the scope of the example embodiments of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrent or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art.

A receiver may have N_RX number of receive antennas. The metric S′(n) in Equation (6) can be calculated by first determining a value of

 P r ⁡ ( n )   R r ⁡ ( n ) 
for each of the ‘N_RX’ receive antennas,
and then averaging those “N_RX” number of

 P r ⁡ ( n )   R r ⁡ ( n ) 
values. To calculate the

 P r ⁡ ( n )   R r ⁡ ( n ) 
value for an r^th receive, antenna after the AGC convergence, 2N_s (e.g., with N_s=16, 32 samples) samples of a received signal are collected at the r^th receive antenna, as shown at block 602. The term N_s represents a length of the STF, for example.

Next, the 2N_s samples are divided into two parts with N_s (e.g., 16) samples in each part, as shown in block 604. For example, a first part has the first N_s samples among the 2N_s samples and the second part has the remaining N_s samples. A conjugate of all the N_s samples of the second part is then taken, as shown at block 606.

Following, a sample wise product of the first N_s samples and the conjugated N_s samples is then obtained, as shown in block 608. Product samples of length N_s will be the result of block 608. Next, an average of the above N_s product samples is calculated, and then a modulus (magnitude) of the average is taken to determine |P_r(n)| at the instant ‘n’, as shown at block 610.

To determine the power estimate, the sample wise product of the second N_s samples and the conjugated N_s samples obtained in block 606 is calculated, as shown in block 612. Product samples of length N_s will be the result. Next, an average of the N_s product samples is calculated, and then a modulus (magnitude) of the average is taken to determine |R_r(n)| at the instant ‘n’, as shown at block 614.

The ratio

 P r ⁡ ( n )   R r ⁡ ( n ) 
is calculated using values obtained in blocks 610 and 614, as shown at block 616. This step will give the value

 P r ⁡ ( n )   R r ⁡ ( n ) 
at the n^th instant, at the r^th receive antenna, by making use of signals received at the r^th receive antenna. Thus, the value

 P r ⁡ ( n )   R r ⁡ ( n ) 
can be calculated for all the “N_RX” receive antennas by making use of receive signals received at respective receive antennas, as shown at blocks 618 and 620.

Following, an average of all the “N_RX”

 P r ⁡ ( n )   R r ⁡ ( n ) 
values found for all N_RX receive antennas is calculated, as shown at block 620, which will provide the metric value S(n) for the instant ‘n’. The metric value S(n) can be found for all the values of ‘n’ up to 9N_s for example. For determining the metric value S(n+1), the 2Ns window is moved by one sample, and the method of the flowchart in FIG. 6 is repeated. Likewise, the metric values S(n+2), S(n+3), S(n+4), . . . can be found. Moreover, these values may not be constant at a maximum value (i.e., due to noise) during the STF portion of the reception. To smoothen the metric values, a weight value is applied to the value found in the past while determining a present metric as shown in Equation (6).

After calculating the metric S′(n) of Equation (6), which may be viewed as a normalized ratio of the autocorrelation value in a received signal to an estimate of the power in the received signal, a second metric (D(n)) can be calculated to be compared to the metric S′(n). The metrics can be used to determine an estimate of the CTO.

The second metric, D(n), is an average power of a difference signal over a window of 2N_s samples (defined from the instant k). The metric is given as:

D ⁡ ( n ) = ⁢ 1 N r ⁢ N s ⁢ ∑ r = 0 N r - 1 ⁢ ⁢ ∑ d = 0 N s - 1 ⁢ ⁢  y r ⁡ ( n + d ) - y r ⁡ ( n + N s + d )  2 ≈ ⁢ 2 ⁢ σ v 2 . Equation ⁢ ⁢ ( 7 )

The value of metric D(n) depends on the instant n. Equation (7) holds true for n<8N_s. For n>8N_s+K with K>0, the metric will be represented as:
D(n)≈2σ_ν²+α(n)+β(n)  Equation (8)
The terms α(n) and β(n) represent averaged power of the STF and LTF respectively. As n increases, the total averaged power of the difference signal will increase when compared to the difference signal in Equation (7). This is because the metric D(n) will have contributions from LTF. So, from the instant n>8N_s, a gradual increase is seen as K increases. A smoothing operation is performed on the metric by weighted averaging, and the new metric is given as:
D′(n)=λD′(n−1)+(1−λ)D(n)  Equation (9)
where λ is a weight factor given to a previous value and (1−λ) is a weight factor applied to the current metric.

Thus, generally, to determine the metric D′(n), the stored samples are indexed and samples separated by an index difference of N_s are subtracted. An absolute value of the resultant is then squared to determine a power of the signal, and a sum of the squared values is averaged over a time window and also over the various receiver antennas, for example.

FIG. 7 is a flowchart that depicts example steps of a method for determining another metric used in a coarse timing offset estimation for a beginning of data within a data packet. Each block may represent a module, a segment, or a portion of program code, which includes one or more instructions executable by a processor for implementing specific logical functions or steps in the process. In addition, each block may represent circuitry that is wired to perform the specific logical functions in the process.

The method of FIG. 7 can be performed in parallel with the method of FIG. 6 to calculated the metrics D(n) and S(n), simultaneously, for example. At each instant of time, the method of FIG. 7 uses the same 2N_s samples used for determining the metric S(n).

To determine the value D_r(n) for the r^th receive antenna, after the AGC convergence has completed, 2N_s (e.g., 32) samples of a received signal are collected at the r^th receive antenna, as shown at block 702. The 2N_s samples are divided into two parts with N_s (16) samples in each part, as shown in block 704. For example, a first includes the first N_s samples among the 2N_s samples, and the second part includes the remaining N_s samples.

Following, a samplewise difference of the first N_s samples and the second N_s samples is calculated, as shown in block 706. Error samples of length N_s will be the result. Because the channel is assumed to be static, the first samples and the second samples are affected by the channel in the same way, so the only difference is noise.

A magnitude square of the error samples is calculated to calculate a power of the noise in the signal, as shown at block 708. Next, an average of the N_s magnitude squared samples is determined to calculated D_r(n) at the instant ‘n’, as shown at block 710.

Block 710 provides the value D_r(n) at n^th instant, at the r^th receive antenna by making use of signals received at r

1. A method for synchronizing wireless transmission of data packets, the method comprising: receiving a plurality of signals, wherein each signal has a data packet frame format including a preamble and data, wherein the preamble comprises a pattern of symbol sequences repeated a number of times; correlating patterns of symbol sequences within the plurality of signals to produce correlation values; calculating a first metric of the plurality of signals, the first metric being a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals; calculating a second metric, the second metric defining an average of a power of noise in the plurality of signals; and determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

2. The method of claim 1, further comprising filtering the first metric through a weight filter to cause values of the first metric to converge as the correlation values decrease.

3. The method of claim 1, further comprising: receiving the plurality of signals at a plurality of receive antennas; calculating the first metric of the plurality of signals for each of the plurality of receive antennas to produce a plurality of first metric values; averaging the plurality of first metric values; calculating the second metric of the plurality of signals for each of the plurality of receive antennas to produce a plurality of second metric values; and averaging the plurality of second metric values.

4. The method of claim 1, wherein N_s samples are included in the pattern of short symbol sequences, and the method further comprises: collecting 2N_s samples of each of the plurality of signal; dividing the 2N_s samples are divided into a first part including a first N_s samples among the 2N_s samples and a second part including the remaining N_s samples; calculating a conjugate of all the N_s samples of the second part; performing a sample wise product of the first N_s samples of the first part and the conjugated N_s samples to produce N_s product samples; calculating an average of the N_s product samples; and calculating a modulus of the average of the N_s product samples to determine the average of the correlation values of the plurality of signals.

5. The method of claim 1, wherein N_s samples are included in the pattern of short symbol sequences, and the method further comprises: collecting 2N_s samples of each of the plurality of signal; dividing the 2N_s samples are divided into a first part including a first N_s samples among the 2N_s samples and a second part including the remaining N_s samples; calculating a conjugate of all the N_s samples of the second part; performing a sample wise product of the second part including the remaining N_s samples and the conjugated N_s samples to produce N_s product samples; calculating an average of the N_s product samples; and calculating a modulus of the average of the N_s product samples to determine the average of the power of the plurality of signals.

6. The method of claim 1, wherein N_s samples are included in the pattern of short symbol sequences, and the method further comprises: collecting 2N_s samples of each of the plurality of signal; dividing the 2N_s samples are divided into a first part including a first N_s samples among the 2N_s samples and a second part including the remaining N_s samples; performing a sample wise difference of the first part including the first N_s samples and the second part including the remaining N_s samples to produce N_s error samples; calculating a magnitude square of each of the N_s error samples; and calculating an average of the N_s magnitude squared samples to determine the average of the power of the noise in the signal.

7. The method of claim 1, further comprising determining an intersection point between the first metric and the second metric as the approximate division of the preamble and the data in the data packet frame format of each signal.

8. The method of claim 7, further comprising determining that the second metric defining the average of the power of noise in the plurality of signals at the intersection point is greater than the second metric defining the average of the power of noise in the plurality of signals at values less than the intersection point; determining that the second metric defining the average of the power of noise in the plurality of signals at the intersection point is less than the second metric defining the average of the power of noise in the plurality of signals at values greater than the intersection point; and verifying that the intersection point between the first metric and the second metric is the approximate division of the preamble and the data in the data packet frame format of each signal.

9. The method of claim 1, further comprising determining an index in the preamble where a sum of a channel impulse response energy between a receive antenna and a transmit antenna is maximum.

10. The method of claim 1, wherein each signal of the plurality of signals includes multiple versions of a transmitted signal, wherein each version is a cyclically shifted version of the transmitted signal, and the method further comprises: determining a cross correlation between the multiple versions of the transmitted signal to produce a number of peaks equal to a number of transmit antennas, each peak corresponding to a total channel energy between a transmit and a receive antenna; and determining an index value needed to substantially shift the number of peaks together, wherein the index corresponds to a fine timing offset amount that defines the approximate division of the preamble and the data in the data packet frame format of each signal.

11. The method of claim 1, wherein each signal of the plurality of signals includes multiple versions of a transmitted signal, wherein each version is a cyclically shifted version of the transmitted signal, and the method further comprises: receiving the plurality of signals at a plurality of receive antennas; at each receive antenna, determining a cross correlation between the multiple versions of the transmitted signal to produce a number of peaks equal to a number of transmit antennas, each peak corresponding to a position to where the transmitted signal was shifted; at each receive antenna, combining cross correlations of the multiple versions of the transmitted signal by shifting correlated outputs so that peaks substantially match to produce combined cross-correlations; adding combined cross-correlations from each antenna to produce a third metric; and identifying a position of a maximum peak within the third metric as a fine timing offset amount that defines the approximate division of the preamble and the data in the data packet frame format of each signal.

12. A computer readable medium having stored therein instructions executable by a computing device to cause the computing device to perform functions of: receiving a plurality of signals, wherein each signal has a data packet frame format including a preamble and data, wherein the preamble comprises a pattern of symbol sequences repeated a number of times; correlating patterns of symbol sequences within the plurality of signals to produce correlation values; calculating a first metric of the plurality of signals, the first metric being a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals; calculating a second metric, the second metric defining an average of a power of noise in the plurality of signals; and determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

13. The computer readable medium of claim 12, wherein the functions further comprise determining an intersection point between the first metric and the second metric as the approximate division of the preamble and the data in the data packet frame format of each signal.

14. The computer readable medium of claim 13, wherein the functions further comprise: determining that the second metric defining the average of the power of noise in the plurality of signals at the intersection point is greater than the second metric defining the average of the power of noise in the plurality of signals at values less than the intersection point; determining that the second metric defining the average of the power of noise in the plurality of signals at the intersection point is less than the second metric defining the average of the power of noise in the plurality of signals at values greater than the intersection point; and verifying that the intersection point between the first metric and the second metric is the approximate division of the preamble and the data in the data packet frame format of each signal.

15. The computer readable medium of claim 12, wherein the functions further comprises determining an index in the preamble where a sum of a channel impulse response energy between a receive antenna and a transmit antenna is maximum.

16. The computer readable medium of claim 12, wherein each signal of the plurality of signals includes multiple versions of a transmitted signal, wherein each version is a cyclically shifted version of the transmitted signal, and the functions further comprise: receiving the plurality of signals at a plurality of receive antennas; at each receive antenna, determining a cross correlation between the multiple versions of the transmitted signal to produce a number of peaks equal to a number of transmit antennas, each peak corresponding to a position to where the transmitted signal was shifted; at each receive antenna, combining cross correlations of the multiple versions of the transmitted signal by shifting correlated outputs so that peaks substantially match to produce combined cross-correlations; adding combined cross-correlations from each antenna to produce a third metric; and identifying a position of a maximum peak within the third metric as a fine timing offset amount that defines the approximate division of the preamble and the data in the data packet frame format of each signal.

17. A system comprising: a processor; a data storage medium; and machine language instructions stored on the data storage medium and executable by the processor to perform the functions of: receiving a plurality of signals, wherein each signal has a data packet frame format including a preamble and data, wherein the preamble comprises a pattern of symbol sequences repeated a number of times; correlating patterns of symbol sequences within the plurality of signals to produce correlation values; calculating a first metric of the plurality of signals, the first metric being a ratio of an average of the correlation values of the plurality of signals and an average of a power of the plurality of signals; calculating a second metric, the second metric defining an average of a power of noise in the plurality of signals; and determining an approximate division of the preamble and the data in the data packet frame format of each signal based on values of the first metric and the second metric.

18. The system of claim 17, wherein each signal of the plurality of signals includes multiple versions of a transmitted signal, wherein each version is a cyclically shifted version of the transmitted signal, and the functions further comprise: receiving the plurality of signals at a plurality of receive antennas; at each receive antenna, determining a cross correlation between the multiple versions of the transmitted signal to produce a number of peaks equal to a number of transmit antennas, each peak corresponding to a position to where the transmitted signal was shifted; at each receive antenna, combining cross correlations of the multiple versions of the transmitted signal by shifting correlated outputs so that peaks substantially match to produce combined cross-correlations; adding combined cross-correlations from each antenna to produce a third metric; and identifying a position of a maximum peak within the third metric as a fine timing offset amount that defines the approximate division of the preamble and the data in the data packet frame format of each signal.

19. The system of claim 17, wherein the system is a multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) receiver.

20. The system of claim 17, further comprising a plurality of receive antennas, each of the plurality of receive antennas receiving one of the plurality of signals, and wherein the functions further comprise: calculating the first metric of the plurality of signals for each of the plurality of receive antennas to produce a plurality of first metric values; and averaging the plurality of first metric values.