Lines Matching defs:rate
234 u8 rate; /* plcp */
248 * High-throughput (HT) rate format for bits 7:0 (bit 8 must be "1"):
264 * Legacy OFDM rate format for bits 7:0 (bit 8 must be "0", bit 9 "0"):
274 * Legacy CCK rate format for bits 7:0 (bit 8 must be "0", bit 9 "1"):
643 * rate tables, and txpower tables. Driver must build a new station
936 * 3945 uses C_RATE_SCALE to set up rate tables).
945 * their own txpower/rate setup data).
1176 u8 rate;
1287 * 1: Use rate scale table (see C_TX_LINK_QUALITY_CMD).
1288 * Tx command's initial_rate_idx indicates first rate to try;
1290 * 0: Use Tx rate/MCS from Tx command's rate_n_flags field.
1291 * This rate will be used for all Tx attempts; it will not be scaled. */
1376 u8 rate;
1424 u8 rate;
1476 * Index into rate table (see C_TX_LINK_QUALITY_CMD) for initial
1479 * rate (via non-0 value) for special frames (e.g. management), while
1480 * still supporting rate scaling for all frames.
1665 * transmitted at same rate. Rate scaling may have been used if first
1706 * member of a previous aggregation block). If rate
1707 * scaling is used, retry count indicates the rate
1763 * rates used for all related commands, including rate
1767 * rate in the rate table, the bit mask for that rate
1789 /* # entries in rate scale table to support Tx retries */
1818 * rs_table[LINK_QUAL_MAX_RETRY_NUM] rate table at which to start.
1885 * The rate scaling procedures described below work well. Of course, other
1891 * Given a particular initial rate and mode, as determined by the rate
1893 * formula to fill the rs_table[LINK_QUAL_MAX_RETRY_NUM] rate table in the
1897 * 1) If using High-throughput (HT) (SISO or MIMO) initial rate:
1898 * a) Use this same initial rate for first 3 entries.
1899 * b) Find next lower available rate using same mode (SISO or MIMO),
1900 * use for next 3 entries. If no lower rate available, switch to
1905 * no MIMO, no short guard interval), at the next lower bit rate
1906 * (e.g. if second HT bit rate was 54, try 48 legacy), and follow
1909 * 2) If using legacy initial rate:
1910 * a) Use the initial rate for only one entry.
1911 * b) For each following entry, reduce the rate to next lower available
1912 * rate, until reaching the lowest available rate.
1913 * c) When reducing rate, also switch antenna selection.
1914 * d) Once lowest available rate is reached, repeat this rate until
1915 * rate table is filled (16 entries), switching antenna each entry.
1920 * The rate scaling algorithm for 4965 devices, as implemented in Linux driver,
1924 * actual transfer rate is better), then the driver continues to use the
1927 * Each history set contains, separately for each possible rate, data for a
1928 * sliding win of the 62 most recent tx attempts at that rate. The data
1937 * attempt might be at a different rate, and have different modulation
1939 * up in the rate scaling table in the Link Quality command. The driver must
1940 * determine which rate table entry was used for each tx attempt, to determine
1941 * which rate-specific history to update, and record only those attempts that
1945 * rate, since there is no per-attempt acknowledgment from the destination
1946 * station. The Tx response struct il_tx_resp indicates the Tx rate in
1954 * driver attempts to find a best initial rate. The initial rate is the
1955 * first entry in the Link Quality command's rate table.
1958 * table below) for current initial rate. Do this only if enough frames
1960 * tx attempts, or at least 8 successes. If not enough, don't try rate
1963 * 2) Find available rates adjacent to current initial rate. Available means:
1972 * 4) Try decreasing rate if, for current rate:
1974 * b) lower adjacent rate has better measured throughput ||
1975 * c) higher adjacent rate has worse throughput, and lower is unmeasured
1977 * As a sanity check, if decrease was determined above, leave rate
1979 * a) lower rate unavailable
1980 * b) success ratio at current rate > 85% (very good)
1982 * of lower rate (under perfect 100% tx conditions, see table below)
1984 * 5) Try increasing rate if, for current rate:
1987 * b) higher adjacent rate has better measured throughput ||
1988 * c) lower adjacent rate has worse throughput, and higher is unmeasured
1990 * As a sanity check, if increase was determined above, leave rate
1992 * a) success ratio at current rate < 70%. This is not particularly
1993 * good performance; higher rate is sure to have poorer success.
1995 * 6) Re-evaluate the rate after each tx frame. If working with block-
2002 * After working with a modulation mode for a "while" (and doing rate scaling),
2020 * When trying a new mode, use the same bit rate as the old/current mode when
2022 * SISO from MIMO or legacy, or to MIMO from SISO or legacy, use a rate
2069 * Rate info; when using rate-scaling, Tx command's initial_rate_idx
2070 * specifies 1st Tx rate attempted, via idx into this table.
2358 * 2) Txpower setting (for rate specified within Tx command)
2504 * Does not include payload. Must specify Tx rate; no rate scaling. */
2548 * Does not include payload. Must specify Tx rate; no rate scaling. */
2692 /* Used for passing to driver number of successes and failures per rate */
3098 * If actual rate of OFDM false alarms (+ plcp_errors) is too high
3101 * each entry. Conversely, if false alarm rate is too low (less than 5
3137 * If actual rate of CCK false alarms (+ plcp_errors) is too high
3156 * If actual rate of CCK false alarms (+ plcp_errors) is too low
3163 * OR 2) 100 or more consecutive beacon periods have had rate of
3176 * If actual rate of CCK false alarms (+ plcp_errors) is within good range