Packet Scheduling in Multipath TCP: Fundamentals, Lessons, and Opportunities

Kimura, Bruno Yuji Lino; Lima, Demetrius C. S. F.; Loureiro, Antônio A. F.

doi:10.1109/jsyst.2020.2965471

Cited by 42 publications

(24 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hence, (35) equals to x i + u opt i y i = u opt i . Since φ > 1 is known a priori, we have from (3) and (5), where, the former one is equivalent to the constraint in ( 26)-( 27) on x i . All these constraints can be written into a unified equation form Az = B with z := (x, y, φ), x := (x 1 , .…”

Section: B Tuning Of Aimd Parametersmentioning

confidence: 99%

“…In particular, scalability, reliability, information sharing/exchange from remote sources are main motivations for the users of distributed systems [4]. For distributed computing systems, many topics have been investigated [5]- [7], such as power management, request scheduling, resource allocation and system reliability, thereby resulting in different models, algorithms and software tools.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimal Resource Scheduling and Allocation under Allowable Over-Scheduling

Ren¹,

Vlahakis²,

Αθανασόπουλος³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper studies optimal scheduling and resource allocation under allowable over-scheduling. Formulating an optimisation problem where over-scheduling is embedded, we derive an optimal solution that can be implemented by means of a new additive increase multiplicative decrease (AIMD) algorithm. After describing the AIMD-like scheduling mechanism as a switching system, we show convergence of the scheme, based on the joint spectral radius of symmetric matrices, and propose two methods for fitting an optimal AIMD tuning to the optimal solution derived. Finally, we demonstrate the overall optimal design strategy via an illustrative example.

show abstract

Section: B Tuning Of Aimd Parametersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Optimal Resource Scheduling and Allocation under Allowable Over-Scheduling

Ren¹,

Vlahakis²,

Αθανασόπουλος³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Multipath (MP) is a feature of transport protocols that enables data exchange between two endpoints, either over different network interfaces or different IP addresses. Selecting the right path to transmit a new packet is not a trivial task, and it can be seen as an entire research line on its own [47]. However, coding techniques can greatly simplify this task, as in MPLOT [48].…”

Section: Coding At Transport Layermentioning

confidence: 99%

Robust QUIC: Integrating Practical Coding in a Low Latency Transport Protocol

et al. 2021

View full text Add to dashboard Cite

We introduce rQUIC, an integration of the QUIC protocol, and a coding module. rQUIC has been designed to feature different coding/decoding schemes and is implemented in go language. We conducted an extensive measurement campaign to provide a thorough characterization of the proposed solution. We compared the performance of rQUIC with that of the original QUIC protocol for different underlying network conditions as well as different traffic patterns. Our results show that rQUIC not only yields a relevant performance gain (shorter delays), especially when network conditions worsen, but also ensures a more predictable behavior. For bulk transfer (long flows), the delay reduction almost reached 70% when the frame error rate was 5%, while under similar conditions, the gain for short flows (web navigation) was ≈ 55%. In the case of the video streaming the QoE gain (p1203 metric) was, approximately, 50%.

show abstract

“…and γi minimizes the right-hand side of (7). Second, from the derivative of ( 7) and the detailed computation, γ−1…”

Section: Optimal Scheduling and Allocationmentioning

confidence: 99%

“…Typically, three types are lightweight computing node (denoted by Node 1) with low service rate and power consumption, middleweight computing node (denoted by Node 2) with medium service rate and power consumption, heavyweight computing node (denoted by Node 3) with high service rate and power consumption. That is, in (4), we set a (7) we can compute that θ 1 = 2.1898, θ 2 = 3.6978, θ 3 = 7.1296. Hence, when the requests come, the lightweight computing node is activated first.…”

Section: Numerical Examplementioning

confidence: 99%

Optimal Resource Scheduling and Allocation in Distributed Computing Systems

Ren¹,

Vlahakis²,

Αθανασόπουλος³

et al. 2021

Preprint

View full text Add to dashboard Cite

The essence of distributed computing systems is how to schedule incoming requests and how to allocate all computing nodes to minimize both time and computation costs. In this paper, we propose a cost-aware optimal scheduling and allocation strategy for distributed computing systems while minimizing the cost function including response time and service cost. First, based on the proposed cost function, we derive the optimal request scheduling policy and the optimal resource allocation policy synchronously. Second, considering the effects of incoming requests on the scheduling policy, the additive increase multiplicative decrease (AIMD) mechanism is implemented to model the relation between the request arrival and scheduling. In particular, the AIMD parameters can be designed such that the derived optimal strategy is still valid. Finally, a numerical example is presented to illustrate the derived results.

show abstract

Packet Scheduling in Multipath TCP: Fundamentals, Lessons, and Opportunities

Cited by 42 publications

References 41 publications

Optimal Resource Scheduling and Allocation under Allowable Over-Scheduling

Optimal Resource Scheduling and Allocation under Allowable Over-Scheduling

Robust QUIC: Integrating Practical Coding in a Low Latency Transport Protocol

Optimal Resource Scheduling and Allocation in Distributed Computing Systems

Contact Info

Product

Resources

About