Operating ranges, tunability and performance of CoDel and PIE

Abstract

Bufferbloat is excessive delay due to the accumulation of packets in a router's oversized queues. CoDel and PIE are two recent Active Queue Management (AQM) algorithms that have been proposed to address bufferbloat by reducing the queuing delay while trying to maintain a high bottleneck utilization. This paper fills a gap by outlining what are the operating ranges, that is the network characteristics (in terms of round-trip times and bottleneck capacity), for which these algorithms achieve their design goals. This new approach to the problem lets us identify deployment scenarios where both AQM schemes result in poor performance when used with default parameters. Because PIE and CoDel have been proposed with RED's deployment issues in mind, it was essential to evaluate to what extent we can tune them to achieve various trade-offs and let them control the queuing delay outside their default operating range. We find that, by appropriate tuning (1) the amount of buffering can easily be controlled with PIE, (2) the Round Trip Time (RTT) sensitivity of CoDel can be reduced. Also, we observe there is more correlation between the congestion level, the achieved queuing delay and the targeted delay with CoDel than with PIE. This paper therefore concludes there is no single overall best AQM scheme, as each scheme proposes a specific trade-off.

Mitigating congestion in wireless sensor networks through clustering and queue assistance: a survey

“…The very attractive nature of CODEL is that it need not to tune with the data rate. It changes the rate dynamically without any impact on utilization (Kuhn et al 2017). But it suffers from issue of overhead of network and not able to look after the flow rate.…”

Adaptive Receiver-Window Adjustment for Delay Reduction in LTE Networks

“…In order to boost the traffic level in the network, the receiver has to intimate the sender to immediately increase the data traffic generation rate in the network based on the rwnd value. e rwnd value is in the range of 101 to 120, and once the sender receives the rwnd value, it does further computation to increase (19) if ((current_time − last_time) > targetperiod) or (counter > targetcounter) then (20) datarate � totaldata/totaldelay (21) currentpercentage � (datarate/datarate max ) × 100% 22Diffpercentage � 100% − currentpercentage (23) if Diffpercentage ≥ limit high then (24) rwnd � (Diffpercentage − limit low ) (25) else if (Diffpercentage ≥ limit low ) then (26) rwnd � (Diffpercentage − limit low ) × (Diffpercentage − limit low /limit high − limit low ) (27) else if (Diffpercentage ≤ lowest) then (28) rwnd � (lowest − Diffpercentage) + 100 (29) end if (30) Avoiding aggressive reduction in congestion window size (31) if ((current_time − last_time) < targetperiod) and (rwnd < 100) then (32) if rwnd > oldrwnd then (33) rwnd � rwnd − oldrwnd (34) oldrwnd � rwnd (35) else (36) oldrwnd � rwnd (37) end if (38) else (39) oldrwnd � 0 (40) end if (41) last_time � current_time (42) totaldata � totaldelay � counter � 0 (43) end if (44) end procedure ALGORITHM 1: Network congestion estimation at the receiver side.…”

“…us, there is overflow of RLC queues due to the large volume of traffic in a short period of time leading to high delay resulting in poor performance. To guarantee high throughput and low delay during congestion, the researchers have proposed various methods such as buffer-aware scheduling [5][6][7][8], active queue management (AQM) techniques [9][10][11][12], receiver window control [13][14][15], loss-based congestion control [16][17][18], delay-based congestion control [16,19,20], rate-based congestion control [4,20,21], admission and congestion control [22][23][24][25], and resource starvation [26][27][28].…”

Software‐defined satellite cloud RAN

“…When a sub-queue is full, incoming packets are dropped. The classifier may consider the nature of the transport protocols in its classification, so that the sub-queues containing flows that are reactive to congestion can consider active queue management [19] techniques to reduce the buffering and the latency [20]. Because they cannot access to the lower layer characteristics, some implementations of network function QoS consider that the whole available goodput can be exploited by the lower layers: This case is referred as the 'clear sky' case.…”