| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 1 | menu "DCCP CCIDs Configuration (EXPERIMENTAL)" | 
| Gerrit Renker | 6d57b43 | 2007-11-21 10:11:52 -0200 | [diff] [blame] | 2 | 	depends on EXPERIMENTAL | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 3 |  | 
| Andrea Bittau | 8d424f6 | 2006-09-19 13:12:44 -0700 | [diff] [blame] | 4 | config IP_DCCP_CCID2_DEBUG | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 5 | 	bool "CCID-2 debugging messages" | 
 | 6 | 	---help--- | 
 | 7 | 	  Enable CCID-2 specific debugging messages. | 
| Gerrit Renker | 8411671 | 2006-11-20 18:26:03 -0200 | [diff] [blame] | 8 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 9 | 	  The debugging output can additionally be toggled by setting the | 
 | 10 | 	  ccid2_debug parameter to 0 or 1. | 
| Andrea Bittau | 8d424f6 | 2006-09-19 13:12:44 -0700 | [diff] [blame] | 11 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 12 | 	  If in doubt, say N. | 
| Andrea Bittau | 8d424f6 | 2006-09-19 13:12:44 -0700 | [diff] [blame] | 13 |  | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 14 | config IP_DCCP_CCID3 | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 15 | 	bool "CCID-3 (TCP-Friendly) (EXPERIMENTAL)" | 
 | 16 | 	def_bool y if (IP_DCCP = y || IP_DCCP = m) | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 17 | 	---help--- | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 18 | 	  CCID-3 denotes TCP-Friendly Rate Control (TFRC), an equation-based | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 19 | 	  rate-controlled congestion control mechanism.  TFRC is designed to | 
 | 20 | 	  be reasonably fair when competing for bandwidth with TCP-like flows, | 
 | 21 | 	  where a flow is "reasonably fair" if its sending rate is generally | 
 | 22 | 	  within a factor of two of the sending rate of a TCP flow under the | 
 | 23 | 	  same conditions.  However, TFRC has a much lower variation of | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 24 | 	  throughput over time compared with TCP, which makes CCID-3 more | 
 | 25 | 	  suitable than CCID-2 for applications such streaming media where a | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 26 | 	  relatively smooth sending rate is of importance. | 
 | 27 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 28 | 	  CCID-3 is further described in RFC 4342, | 
| Gerrit Renker | 0e64e94 | 2006-10-24 16:17:51 -0700 | [diff] [blame] | 29 | 	  http://www.ietf.org/rfc/rfc4342.txt | 
| Andrea Bittau | 2a91aa3 | 2006-03-20 17:41:47 -0800 | [diff] [blame] | 30 |  | 
 | 31 | 	  The TFRC congestion control algorithms were initially described in | 
| Gerrit Renker | 4dbc242 | 2009-01-11 00:17:22 -0800 | [diff] [blame] | 32 | 	  RFC 5348. | 
| Andrea Bittau | 2a91aa3 | 2006-03-20 17:41:47 -0800 | [diff] [blame] | 33 |  | 
| Gerrit Renker | 0e64e94 | 2006-10-24 16:17:51 -0700 | [diff] [blame] | 34 | 	  This text was extracted from RFC 4340 (sec. 10.2), | 
 | 35 | 	  http://www.ietf.org/rfc/rfc4340.txt | 
| Gerrit Renker | 8411671 | 2006-11-20 18:26:03 -0200 | [diff] [blame] | 36 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 37 | 	  If in doubt, say N. | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 38 |  | 
| Gerrit Renker | 56724aa | 2006-11-20 18:28:09 -0200 | [diff] [blame] | 39 | config IP_DCCP_CCID3_DEBUG | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 40 | 	bool "CCID-3 debugging messages" | 
 | 41 | 	depends on IP_DCCP_CCID3 | 
 | 42 | 	---help--- | 
 | 43 | 	  Enable CCID-3 specific debugging messages. | 
| Gerrit Renker | 56724aa | 2006-11-20 18:28:09 -0200 | [diff] [blame] | 44 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 45 | 	  The debugging output can additionally be toggled by setting the | 
 | 46 | 	  ccid3_debug parameter to 0 or 1. | 
| Gerrit Renker | 56724aa | 2006-11-20 18:28:09 -0200 | [diff] [blame] | 47 |  | 
| Gerrit Renker | ddebc97 | 2009-01-04 21:42:53 -0800 | [diff] [blame] | 48 | 	  If in doubt, say N. | 
| Gerrit Renker | 8a508ac | 2006-12-03 14:50:23 -0200 | [diff] [blame] | 49 |  | 
 | 50 | config IP_DCCP_CCID3_RTO | 
 | 51 | 	  int "Use higher bound for nofeedback timer" | 
 | 52 | 	  default 100 | 
 | 53 | 	  depends on IP_DCCP_CCID3 && EXPERIMENTAL | 
 | 54 | 	  ---help--- | 
 | 55 | 	    Use higher lower bound for nofeedback timer expiration. | 
 | 56 |  | 
 | 57 | 	    The TFRC nofeedback timer normally expires after the maximum of 4 | 
 | 58 | 	    RTTs and twice the current send interval (RFC 3448, 4.3). On LANs | 
 | 59 | 	    with a small RTT this can mean a high processing load and reduced | 
 | 60 | 	    performance, since then the nofeedback timer is triggered very | 
 | 61 | 	    frequently. | 
 | 62 |  | 
 | 63 | 	    This option enables to set a higher lower bound for the nofeedback | 
 | 64 | 	    value. Values in units of milliseconds can be set here. | 
 | 65 |  | 
 | 66 | 	    A value of 0 disables this feature by enforcing the value specified | 
 | 67 | 	    in RFC 3448. The following values have been suggested as bounds for | 
 | 68 | 	    experimental use: | 
| Gerrit Renker | aa1b1ff | 2009-09-12 07:47:01 +0000 | [diff] [blame] | 69 | 		* 16-20ms to match the typical multimedia inter-frame interval | 
 | 70 | 		* 100ms as a reasonable compromise [default] | 
 | 71 | 		* 1000ms corresponds to the lower TCP RTO bound (RFC 2988, 2.4) | 
| Gerrit Renker | 8a508ac | 2006-12-03 14:50:23 -0200 | [diff] [blame] | 72 |  | 
 | 73 | 	    The default of 100ms is a compromise between a large value for | 
 | 74 | 	    efficient DCCP implementations, and a small value to avoid disrupting | 
 | 75 | 	    the network in times of congestion. | 
 | 76 |  | 
 | 77 | 	    The purpose of the nofeedback timer is to slow DCCP down when there | 
 | 78 | 	    is serious network congestion: experimenting with larger values should | 
 | 79 | 	    therefore not be performed on WANs. | 
 | 80 |  | 
| Gerrit Renker | c40616c | 2007-12-06 12:26:38 -0200 | [diff] [blame] | 81 | config IP_DCCP_TFRC_LIB | 
| Gerrit Renker | 129fa44 | 2009-01-04 21:45:33 -0800 | [diff] [blame] | 82 | 	def_bool y if IP_DCCP_CCID3 | 
| Gerrit Renker | c40616c | 2007-12-06 12:26:38 -0200 | [diff] [blame] | 83 |  | 
 | 84 | config IP_DCCP_TFRC_DEBUG | 
| Gerrit Renker | 129fa44 | 2009-01-04 21:45:33 -0800 | [diff] [blame] | 85 | 	def_bool y if IP_DCCP_CCID3_DEBUG | 
| Arnaldo Carvalho de Melo | 7c65787 | 2005-08-09 20:14:34 -0700 | [diff] [blame] | 86 | endmenu |