Changes between Version 29 and Version 30 of OFDMReferenceDesign/Changelog

Timestamp:

Apr 9, 2009, 11:48:34 PM (15 years ago)

Author:

murphpo

Comment:

--

OFDMReferenceDesign/Changelog

@@ -11 +11 @@
  * Packet detector updates
    * Integrated the RSSI-based packet detector core into the OFDM transceiver.
+   * Removed the I/Q-based packet detector that was previously present. This is now only used in simulation.
    * The integrated detector's registers were rearranged (relative to the older separate detector core)
    * Added a minimum duration requirement to the packet detector. This block requires the average RSSI exceed the programmed energy threshold for a minimum number of cycles before the packet detecting output is asserted. The required minimum duration is programmable via a register.
@@ -21 +22 @@
  * Upgraded EEPROM controller core HDL with better clocking design. Instead of generating a slow clock, the core generates a slow clock enable and uses the fast clock for all synchronous elements. This fixes the long-standing, intermittent timing error.
  * Fixed the PHY's handling of packet errors when all received bytes are zero. Previously this caused an erroneous good packet interrupt; now it correctly asserts the bad header and bad packet status bits.
+ * Added a programmable saturating scale factor after the IFFT in the transmitter. This can increase the average transmission power at the cost of increased clipping. This scale factor has been defaulted to a value that jointly maximizes transmission power while minimizing transmitted EVM.
 
 '''Software Changes'''
  * Updated WARPPHY and WARPMAC with register map for integrated transceiver and packet detector.
  * Updated EEPROM driver to match new hardware.
+ * Replaced warpmac_pollEthernet with the more general warpmac_pollPeripherals
+   * This function polls the timer, the PHY, and the Ethernet
+   * One of the four timers in the warp_timer core is dedicated to the polling of the UART and User I/O buttons
+ * Fixed bug in the function for setting random backoff timers found by forum user domenique (See post [http://warp.rice.edu/forums/viewtopic.php?id=465 here]). Previously the minimum contention window was ![0,7]. This is now ![0,15] like 802.11
+ * Lowered packet detection threshold from 9000 to 3400. This should improve low SNR detection
+ * Raised CSMA threshold from 6000 to 8192. This is more in line with 802.11 and should make nodes more aggressive in the presence of non-decodable interference 
+
+
+
+