3 Reviews * Review 1 (Reviewer A) Familiarity Expert (I conduct(ed) active research work in this topic) (4) Recommendation accept: top 20% of all submissions, but not top 10% (4) Contributions (What are the major issues addressed in the paper? Do you consider them important? Comment on the novelty, creativity, impact, and technical depth in the paper.) The paper proposes a novel way to use heterogeneous channels in a multi-channel network. The idea is to set different parameters for different channels (and in this case these parameters are CS thresholds) and allow nodes to choose different channels based on their channel state, which in turn improves network performance. Strengths (What are the major reasons to accept the paper? [Be brief.]) The idea is simple, cute and novel. I haven’t seen it before and I quite like it. Weaknesses (What are the major reasons NOT to accept the paper? [Be brief.]) The evaluation is limited. Detailed Comments (Please provide detailed comments that will help the TPC assess the paper and help provide feedback to the authors.) I really like the idea. It is incredibly simple and it seems to work. But the evaluation is limited. The authors do not offer much insight in why it works. They explicitly say that it is out of scope of the paper to find the best combination of CS thresholds. Why? The paper would have been much nicer if they have provided some insight through an analysis, which would probably also tell us how to choose these values. Right now it is cute and half baked. * Review 2 (Reviewer C) Familiarity Familiar (I am well aware of research work in this topic) (3) Recommendation accept: top 20% of all submissions, but not top 10% (4) Contributions (What are the major issues addressed in the paper? Do you consider them important? Comment on the novelty, creativity, impact, and technical depth in the paper.) This paper presents a dynamic channel switching scheme in WLAN considering the heterogeneous channel conditions. The problem is similar to the traditional rate adaptation in WLAN. In this paper, each link has a target rate and once the rate cannot be achieved, the source node of the link will lower the target rate. However, in this paper, the authors introduces a new option of switching the channel before lowering the target rate. The solution is evaluated with NS3. Strengths (What are the major reasons to accept the paper? [Be brief.]) - the paper is well organized and clearly written - the evaluation is quite comprehensive Weaknesses (What are the major reasons NOT to accept the paper? [Be brief.]) - the algorithm itself is straightforward Detailed Comments (Please provide detailed comments that will help the TPC assess the paper and help provide feedback to the authors.) Overall, it's a quite solid system paper. The authors consider the heterogeneous conditions of different channels. Each link is specified with a target rate. Once the rate cannot be achieved, the node will first try to find another alternative channel for the target rate. If there's no candidate channel found, the node will lower the target rate. The proposed algorithm is straightforward. But the authors implemented the solution on NS3 simulator and conducted extensive evaluation (the evaluation part takes more than half of the paper). I feel the results are supportive and valuable. From the design's perspective, my only concern is about the assumption of the APs. I feel one of the big challenges for any channel switching/hopping scheme is how to make both nodes cooperatively change to the same channel. This paper simplified the problem by assuming that the APs are able to monitor all the channels and upon receiving the packets from a node in the new channel, the associated AP detects the change. * Review 3 (Reviewer B) Familiarity Some knowledge (I am marginally aware of research work in this topic) (2) Recommendation accept: top 20% of all submissions, but not top 10% (4) Contributions (What are the major issues addressed in the paper? Do you consider them important? Comment on the novelty, creativity, impact, and technical depth in the paper.) This paper proposes a scheme to improve performance in a wireless system - a dynamic channel switching protocol along with the idea of differing carrier sensing thresholds across channels. The paper doesn't propose a specific scheme for determining the carrier sensing thresholds but their evaluation studies performance across a range of values (via NS3 simulations). Strengths (What are the major reasons to accept the paper? [Be brief.]) - Simulations show promising results (up to 26% improvement in medium throughput compared to a "homogenous" channel - i.e. same Csr) Weaknesses (What are the major reasons NOT to accept the paper? [Be brief.]) - An important part of the story is left for the future - how to pick the right carrier sensing thresholds? - Work is applicable mainly in specific scenarios - enterprise wifi comes to mind. In a dense apartment complex with a large number of competing wifi systems, I doubt this helps (maybe even makes things worse). - Dynamic channel switching isn't a new idea Detailed Comments (Please provide detailed comments that will help the TPC assess the paper and help provide feedback to the authors.) I'm not an expert in wireless but I was able to understand the core idea proposed in the paper - if you have heterogeneous channels (e.g. when carrier sensing for radio interference thresholds are different), dynamic channel switching can be beneficial. Essentially, "good" links (low interference, good signal strength) can be assigned to channels with low carrier sensing thresholds. The opportunistic channel switching algorithm proposed in this paper is simple - for a given link, it is preferred to switch channels before lowering data rates. The paper doesn't propose a specific scheme for determining the carrier sensing thresholds. It is claimed to be out of scope of this paper but I'm not sure I see it that way. The evaluation attempts to make up for this by studying performance across a range of Csr values. The simulation studies presented are promising - in the best case, 26% improvement in medium throughput is observed compared to a "homogenous" channel - i.e. same Csr. Overall, the paper has some nice ideas but leaves unanswered questions.