Adapting TCP Small Queues for IEEE 802.11 Networks

Grazia, Carlo Augusto; Patriciello, Natale; Høiland-Jørgensen, Toke; Klapez, Martin; Casoni, Maurizio; Mangues‐Bafalluy, Josep

doi:10.1109/pimrc.2018.8581048

Cited by 14 publications

(23 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The standard TSQ behavior on wired networks is to allow each TCP socket to enqueue a number of packets that is equivalent to the number of packets that would be sent in 1 ms at the current sending rate; this mechanism helps in maintaining an upper bound of the queueing delay of the sending node as a function of the flow throughput. This global constraint of 1 ms has been proved in [11] to be too strict in a Wi-Fi environment where the frame aggregation is not possible with such a limit.…”

Section: Stackmentioning

confidence: 99%

“…Even if the TSQ performance over wired links is remarkable, this is not the case for WLAN environments in which TSQ could break the frame aggregation logic, impeding all the TCP variants to discover the full link potential correctly. This limitation has been discovered and solved for TCP Cubic, and it led to a new solution for boosting BBR v2.0 throughput on Wi-Fi paths [11]- [13]. Unfortunately, applying the same fix to the BBR algorithm is not enough to get a decent throughput from the Wi-Fi interface.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

BBRp: Improving TCP BBR Performance Over WLAN

2020

Self Cite

View full text Add to dashboard Cite

This paper shows the inefficiency of TCP BBR in exploiting the Wi-Fi bandwidth. This limitation of BBR has been observed with both IEEE 802.11n and IEEE 802.11ac, where the mechanism of frame aggregation is used to boost the throughput of data transmission. In the last years, many TCP variants have been introduced to limit the bufferbloat phenomena and bound the latency through a reduction of the queue backlog injection rate. However, this mechanism impacts on the Wi-Fi frame aggregation logic, impeding TCP congestion controls to reach the full throughput potential of a Wi-Fi interface. While this problem can be solved with TCP Cubic by allowing the sender node to enqueue more packets, for TCP BBR the fix is not the same, as it has a customized pacing algorithm. With this contribution we propose BBRp, a new BBR version that allows for fine-tuning the congestion control pace, achieving between four and six times more throughput over IEEE 802.11n and IEEE 802.11ac channels, at the cost of an increased latency that is however always less than the latency obtainable with loss-based TCP congestion controls.

show abstract

Section: Stackmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

BBRp: Improving TCP BBR Performance Over WLAN

2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…The BQL algorithm resizes the queue to its optimal byte size according to the last egress rate. Such a dynamical mechanism has proved to be adequate since a static small queue may reject packets before achieving the full transmission rate capability [30]. Even though good simulation results were obtained, the study considers the cellular access network as a queue, while in reality, the QoS queuing is composed of multiple hierarchical queues [1] (c.f.…”

Section: Tackling the Bufferbloat Phenomena From The Queues Perspectivementioning

confidence: 99%

“…In these experiments both low latency flows were generated with a normal distribution and a 5 ms standard deviation. One was created with an average interval time of 50 ms, while the average interval for the second one was tested for [20,30,40,50,60,70] ms. The results obtained are very similar to the ones presented in this section quantitatively.…”

Section: Shared Sdap Buffer Shared Rlc Buffermentioning

confidence: 99%

Dynamic Buffer Sizing and Pacing as Enablers of 5G Low-Latency Services

Irazabal

López-Aguilera

Demirkol

et al. 2022

IEEE Trans. on Mobile Comput.

View full text Add to dashboard Cite

3GPP standards organization is performing an impressive effort trying to reach sub-millisecond latencies for 5G. However, such efforts may become fruitless if exogenously generated delays at transport layer are not considered. Nowadays, Radio Access Networks (RANs) are deployed with large buffers to achieve full utilization and avoid squandering wireless resources. Unfortunately, and since the data path's bottleneck resides on the radio link, RAN's buffers are bloated by TCP's congestion control algorithm. Thus, a flow with low-latency requirements that encounters a bloated buffer, suffers from inevitable large sojourn times associated with the buffer depletion time, severely downgrading its Quality of Service (QoS). This paper presents different solutions for efficiently multiplexing distinct traffic patterns that share buffers on the 5G stack. Bufferbloat is extensively studied within the actual 5G QoS scenario, which presents multiple challenges inherited from the dynamic radio link nature and the presence of multiple queues at different entities. We propose and extensively emulate different algorithms in order to avoid the exogenous delay caused by the bufferbloat phenomena. We use real cellular network traces with realistic delay-sensitive and background traffic patterns in different scenarios. The outcome presents valuable insights in the algorithms that will enable low-latency services to be delivered through the 5G network stack satisfying restrictive envisioned constraints.

show abstract

“…IEEE 802.11 increases its probability of successfully transmitting a packet in a noisy channel through packet fragmentation. However, the asynchronous nature of accessing the channel, compels to aggregate frames for achieving full bandwidth [41], and therefore, the fragmentation procedure is rare in comparison with the RLC segmentation/reassembly mechanism that arises in cellular networks due to their synchronous nature. Therefore, the segmentation/reassembly procedure can be mostly considered as a cellular network specificity, and consequently, it has not been thoroughly researched in IEEE 802.3 and IEEE 802.11 standards.…”

Section: ) Problem Descriptionmentioning

confidence: 99%

Preventing RLC Buffer Sojourn Delays in 5G

et al. 2021

View full text Add to dashboard Cite

The 3rd Generation Partnership Project (3GPP) is investing a notable effort to mitigate the endogenous stack and protocol delays (e.g., introducing new numerology, through preemptive scheduling or providing uplink granted free transmission) to attain to the heterogeneous Quality of Service (QoS) latency requirements for which the fifth generation technology standard for broadband cellular networks (5G) is envisioned. However, 3GPP's goals may become futile if exogenous delays generated by the transport layer (e.g., bufferbloat) and the Radio Link Control (RLC) sublayer segmentation/reassembly procedure are not targeted. On the one hand, the bufferbloat specifically occurs at the Radio Access Network (RAN) since the data path bottleneck is located at the radio link, and contemporary RANs are deployed with large buffers to avoid squandering scarce wireless resources. On the other hand, a Resource Block (RB) scheduling that dismisses 5G's packet-switched network nature, unnecessarily triggers the segmentation procedure at sender's RLC sublayer, which adds extra delay as receiver's RLC sublayer cannot forward the packets to higher sublayers until they are reassembled. Consequently, the exogenously generated queuing delays can surpass 5G's stack and protocol endogenous delays, neutralizing 3GPP's attempt to reduce the latency. We address RLC's related buffer delays and present two solutions: (i) we enhance the 3GPP standard and propose a bufferbloat avoidance algorithm, and (ii) we propose a RB scheduler for circumventing the added sojourn time caused by the packet segmentation/reassembly procedure. Both solutions are implemented and extensively evaluated along with other state-of-the-art proposals in a testbed to verify their suitability and effectiveness under realistic conditions of use (i.e., by considering Modulation and Coding Scheme (MCS) variations, slices, different traffic patterns and off-the-shelf equipment). The results reveal current 3GPP deficits in its QoS model to address the bufferbloat and the contribution of the segmentation/reassembly procedure to the total delay.

show abstract

Adapting TCP Small Queues for IEEE 802.11 Networks

Cited by 14 publications

References 13 publications

BBRp: Improving TCP BBR Performance Over WLAN

BBRp: Improving TCP BBR Performance Over WLAN

Dynamic Buffer Sizing and Pacing as Enablers of 5G Low-Latency Services

Preventing RLC Buffer Sojourn Delays in 5G

Contact Info

Product

Resources

About