From: Salil S Kanhere Subject: [LCN 2015] Your paper #1570138741 ('O-ACK: An Adaptive Wireless MAC Protocol Exploiting Opportunistic Token-Passing and Ack Piggybacking') Date: July 5, 2015 5:30:40 PM CDT To: Shegufta Ahsan , Nitin Vaidya Cc: Salil S Kanhere , Nils Aschenbruck , Jens Toelle , "Matthias Hollick" Dear Shegufta Ahsan: Congratulations - we are pleased to inform you that your paper #1570138741, = entitled 'O-ACK: An Adaptive Wireless MAC Protocol Exploiting Opportunistic = Token-Passing and Ack Piggybacking', has been accepted in the "Short_Papers" = track at 40th Annual IEEE Conference on Local Computer Networks and for = publication in its proceedings. This means your paper is accepted and will be published in the proceedings, = but it needs to be cut to 4 pages and a poster has to be prepared for the = presentation. Please revise your paper carefully to address the reviewers' comments and = suggestions and to ensure that your final paper fits the camera-ready = format. The reviews, attached at the end of this email, are also available at = http://edas.info/showPaper.php?m=1570138741. At least one author must register at full author rate for the conference by = July 30, 2015. For information on registration, visa, venue, etc. please visit: http://www.ieeelcn.org/Attendee.html The camera ready copy of your paper is due July 30, 2015, via EDAS. Instructions on final manuscript preparation as well as IEEE copyright = transfer will be available soon at: = http://ieeelcn.org/Author_Instructions_Final.html Please make sure to follow all instructions. Authors are strongly encouraged to submit a proposal for the *Demo* Session = at LCN. This session presents an excellent opportunity for authors to = showcase a prototype implementation of the research ideas presented in the = paper. Moreover, authors of accepted papers do NOT need to pay additional = registration fees if the demo is accepted for presentation. Demo proposals are due July 15, 2015. For further details, please visit: http://www.ieeelcn.org/Author_CfD.html A limited number of *student grants* are available to attend the LCN 2015 = conference. The applicant must be a student author of an accepted = full/poster paper of the main conference or a student author of an accepted = workshop paper. The grants are sponsored by the IEEE Computer Society = Technical Committee on Computer Communications (TCCC). Each grant will = cover one free author registration to the LCN 2015 conference, for one = paper. Submissions of application are due July 24, 2015. For further information please visit: = http://www.ieeelcn.org/Author_Student.html We look forward to seeing you at LCN in October! Kind regards, Salil Kanhere, TPC Chair Jens Toelle, TPC Co-chair ======================== ======= Review 1 ======= *** Strengths: What are the most important merits of the paper?  Please = explain in one or more sentences. The authors present a novel MAC protocol “O-ACK” for = wireless LAN, with the goal of reducing channel idle time and collision = overhead while still allowing for newly arriving stations and adjacent APs = using the same channel. *** Weaknesses: What are the major weaknesses in the paper? Please explain = in one or more sentences. Difficult to read, explanations in design and regarding evaluation results = are missing. No comparison to related work other that a list of existing approaches. Evaluation against plain 802.11, contribution against state-of-the-art = unclear. The basic idea is commendable, but in the current form, the paper has too = many weaknesses, see “Feedback” section. *** Feedback to Authors: Please provide detailed comments that will help = the authors to understand the weaknesses in their paper and improve their = work.  This section is very important if you rate the paper as a = reject. Related Work: - While the authors list related approaches, no information is given as to = why the state-of-the-art requires another paper and how O-ACK improves upon = the existing works. - Especially in the face of such strong “competition” = (CONEXT, MOBICOM, INFOCOM, ToN) and even an IEEE 802.11 standard, a clear = discussion towards the merits and differences of the new protocol is = necessary. - Where an explanation is given (“extra overhead”, = “conflict graph of network links remains same over time”, = “must have at least two packets destined for the same = receiver”, “optimized only for down-link traffic”, = etc), these are not picked up in either the design or the evaluation. - The similarities between O-ACK and related works extend to specific = wording, such as “clients automatically start transmitting = chains” and “multiple frames separated by a SIFS = interval”, making it difficult to judge the novelty without an = elaboration on the differences. - Please provide a reference for Barcelo et al. (page 2, 2nd column). Design : - The idea of ACK piggybacking in 802.11 to reduce the idle time from DIFS = to SIFS is not new and has already been proposed in other papers, as = highlighted by the amount of referenced related work. - It is not clear if legacy devices, not implementing O-ACK, may still = associate to an O-ACK AP or, if the lambda parameter allows this, how their = performance is in an O-ACK network in terms of fairness. - This is an important information when thinking of future deployment = scenarios of O-ACK. - The ability to overhear piggybacked ACKs (or data) requires stations to = listen to and process *every* packet sent from the AP (since the bit-wise = check of their MAC address and the network identifier does no longer = suffice), inducing massive overhead at the station. This is especially = true if, as stated, the throughput in the network needs to be increased = beyond the maximum afforded by 802.11 DCF. For battery-driven devices = (smartphones, laptops), that make up the vast majority of wireless devices, = this overhead is harmful towards their battery lifetime (see also 802.11 = monitor mode). The authors should at least discuss this. - The discussion of multiple APs on the same channel is unclear. 802.11 STAs = filter frames by the network identifier, hence why would the presence of = multiple APs lead to corrupted packets ? Of course, the channel capacity is = limited by concurrent transmissions as in CSMA-CA, but frames of other APs = or foreign STAs should not be received if AP and STA are not running in = Monitor mode. In this, how do APs overhear frames from STAs connected to = foreign APs to construct the set STA_overheard? - How does setting the current_max_prob probability ensure fair use of the = channel ? The paper does not mention any coordination between APs and no = correlation between the traffic requests of STAs in foreign networks. As = such, it seems that collisions will still occur with the same probability as = in 802.11 DCF. - Finally, how does the assignment of privileged stations work? The meaning = of “with probability λ, AP selects a station as privileged = station” is unclear. Which station is selected with this = probability ? A random one ? The benefit of assigning a privileged station = very much hinges on whether that station has data to transmit, which does = not seem possible for the AP to determine, otherwise the assignment is = overhead. Evaluation : - The whole evaluation chapter is in one piece, making it extremely = difficult to read and understand as well as to keep track of the evaluated = design aspects. - Also the traffic load used in the simulations should be indicated in Table = I and simulation results for different traffic loads should be provided. - Please indicate for each scenario how many simulation runs you = performed. - Until the second evaluation scenario (“dense”), no = explanation of the simulated scenario is given. I.e. how much/when do = stations send, does every station always have data to send, is there station = mobility? - Most importantly, the evaluation does not reveal how O-ACK performs under = different traffic scenarios and thus if O-ACK always outperforms standard = DCF. - The setting of four stations joining the network in bulk is unrealistic = and the results are not explained. Why does the evaluation target = throughput instead of the time it takes to join the network with regard to = lambda ? - Do the joining stations all join AP1 or are they spread over multiple APs = ? This would have tremendous impact on both the throughput and the joining = time. - Fig. 3 as well as the explanation leading to Figure 4 do not seem to serve = a purpose. - The sequence of Figures 4-6 does not match the sequence in which they are = addressed in the evaluation. - In general, the results are presented without explanation and reference to = the 802.11 DCF base line. Again, this could be mitigated by explaining = the evaluation scenario and parameters to give context to the numbers. - A good starting point for such context would be a systematic = identification of collision overhead and channel idle time in 802.11 DCF = and O-ACK, as in the abstract and introduction these are claimed to be the = most important drawbacks. - Why is the comparison between 802.11g and O-ACK ? Do features such as = Channel Bonding, reduced guard times, etc. in 802.11n (positively or = negatively) impact the design and performance of O-ACK? - Very important : Regarding the amount of high-quality related work, a = comparison between O-ACK and the existing approaches would be more = interesting than between O-ACK and plain 802.11. Otherwise it is not = possible to positively assess the contribution. *** Scientific/Technical Quality: How would you rate the = scientific/technical quality of the paper? (1 being worst score, 5 being = best score) Marginal (2) *** Innovation: How would you rate the innovation of the paper, the news = being received by the reader? (1 being worst score, 5 being best score) Marginal (2) *** Presentation Quality: How would you rate the quality of presentation = within the paper? (1 being worst score, 5 being best score) Marginal (2) *** Overall Recommendation: What is your over all recommendation for this paper? Reject (2) ======= Review 2 ======= *** Strengths: What are the most important merits of the paper?  Please = explain in one or more sentences. Authors present a MAC protocol which improves the channel utilization by = leveraging piggybacking, packet overhearing, token based scheduling. The protocol is also able to adjust itself taking into = account the current channel conditions. *** Weaknesses: What are the major weaknesses in the paper? Please explain = in one or more sentences. The packet overhearing is a fundamental element of the protocol but this = technique is not always usable. The proposal does not take into account the node mobility and only one kind = of traffic source is considered. Simulation are performed in quite static conditions and, for example, it is = difficult to predict what would happen by varying the node's  turnover. *** Feedback to Authors: Please provide detailed comments that will help = the authors to understand the weaknesses in their paper and improve their = work.  This section is very important if you rate the paper as a = reject. Performance evaluation is performed in quite simplified conditions, = therefore it is difficult to assess the real usefulness of the proposed = protocol. Authors should provide a performance evaluation in presence of mobile nodes = and heterogeneous traffic sources. *** Scientific/Technical Quality: How would you rate the = scientific/technical quality of the paper? (1 being worst score, 5 being = best score) Excellent (4) *** Innovation: How would you rate the innovation of the paper, the news = being received by the reader? (1 being worst score, 5 being best score) Average (3) *** Presentation Quality: How would you rate the quality of presentation = within the paper? (1 being worst score, 5 being best score) Excellent (4) *** Overall Recommendation: What is your over all recommendation for this paper? Accept (4) ======= Review 3 ======= *** Strengths: What are the most important merits of the paper?  Please = explain in one or more sentences. Authors propose an ACK handling and token mechanism for 802.11 MAC to = improve the throughput performance of the default MAC protocol. The ideas of using tokens and ACK piggybacking are potential methods to = reduce the collisions and channel idle times, respectively. The possible = starvation problem encountered in token-based methods is also tackled by = capping the token-based channel access. The ns-2 simulation evaluations of three network setups show that the = proposed approach can improve the default IEEE 802.11 and preliminary = version of the protocol (T-DCF) and reduces the retransmission ratio. *** Weaknesses: What are the major weaknesses in the paper? Please explain = in one or more sentences. The selection method of privileged node is not clearly defined. Assuming it = is done randomly or in a round-robin fashion, in a dynamic network where the = links are transient and traffic is bursty, the advantage of this approach is = expected to be limited. However, although it is common, such type of = scenarios are not evaluated in detail in the paper.   The token-based medium access is similar to polling, and hence 802.11 PCF. = In relatively static networks (as studied in this paper), PCF is expected = provide a similar maximum throughput performance to T-DCF and potentially to = O-ACK.  In this type of setup, the quantitative advantages of the O-ACK is = questionable. The random flow setup does not state the distribution used for the flow = duration. The HCF (hybrid coordination function) controlled channel access (HCCA) = defined in 802.11e also  give priority to one station over another, = supports Block ACK to reduce ACK overhead. The contribution of O-ACK in = comparison to HCCA is not clear. *** Feedback to Authors: Please provide detailed comments that will help = the authors to understand the weaknesses in their paper and improve their = work.  This section is very important if you rate the paper as a = reject. Per-flow throughput performance depicted in Fig. 10 depends on the number of = other concurrent flows (randomly started and stopped) and the random choice = of the privileged node. Hence, per-flow comparison does not yield = scientifically-significant detail. More realistic scenarios (intermittent data flows, STAs joining and leaving, = having link errors, etc.) are important to provide average behavior of = O-ACK. ACK piggybacking and token-based medium access ideas are not novel, = however, evaluation of them in the widely adopted 802.11 standard is = interesting. *** Scientific/Technical Quality: How would you rate the = scientific/technical quality of the paper? (1 being worst score, 5 being = best score) Average (3) *** Innovation: How would you rate the innovation of the paper, the news = being received by the reader? (1 being worst score, 5 being best score) Average (3) *** Presentation Quality: How would you rate the quality of presentation = within the paper? (1 being worst score, 5 being best score) Excellent (4) *** Overall Recommendation: What is your over all recommendation for this paper? Accept if Room (3) ======= Review 4 ======= *** Strengths: What are the most important merits of the paper?  Please = explain in one or more sentences. Very well referenced, fairly well written, reasonably well presented, and = thoroughly evaluated (with simulation). *** Weaknesses: What are the major weaknesses in the paper? Please explain = in one or more sentences. The motivation for this work is not entirely clear. The contribution seems = rather dated. *** Feedback to Authors: Please provide detailed comments that will help = the authors to understand the weaknesses in their paper and improve their = work.  This section is very important if you rate the paper as a = reject. On the whole, a very good paper that deserves publication. The technical = content is fine; even though I'm not a big fan of simulation-oriented work = in this space, the simulation-oriented evaluation in this paper is fairly = extensive and does a good job of exploring the parameter space. I also = appreciate the thoroughness of the protocol description and the very rich = related work section. The key problems are: (1) the motivation is not made clear; (2) the = timeliness of the contribution is questionable. Indeed, one look through = the set of references shows there's not much recent work in this space the = authors care to refer to. On the whole, the paper deserves publication, but the authors should take = the time to address (1) and ask themselves whether I have a point about = (2). *** Scientific/Technical Quality: How would you rate the = scientific/technical quality of the paper? (1 being worst score, 5 being = best score) Excellent (4) *** Innovation: How would you rate the innovation of the paper, the news = being received by the reader? (1 being worst score, 5 being best score) Average (3) *** Presentation Quality: How would you rate the quality of presentation = within the paper? (1 being worst score, 5 being best score) Excellent (4) *** Overall Recommendation: What is your over all recommendation for this paper? Accept (4) ======= Review 5 ======= *** Strengths: What are the most important merits of the paper?  Please = explain in one or more sentences. This paper proposed a new MAC protocol which improves throughput by = minimizing DIFS via packet overhearing and combined ACK and DATA packets. = It takes advantage of the fact that sender and receiver of a node i can = overhead the packets i transmits. Additional mechanisms to improve = bandwidth utilization (token-based scheduling) and avoid sarvation = (condition-based adapatation) are also incorporated. *** Weaknesses: What are the major weaknesses in the paper? Please explain = in one or more sentences. 1. How does this mechanism perform under hidden and exposed terminal = scenarios? These scenarios occur more frequent in ad-hoc networks. 2. Fairness index should be included in the performance evaluation. 3. With the proposed method, collisions in hidden terminal scenarios might = be further aggravated in ad-hoc networks. 4. More specific plans to extend this work to support ad-hoc network shall = be provided. *** Feedback to Authors: Please provide detailed comments that will help = the authors to understand the weaknesses in their paper and improve their = work.  This section is very important if you rate the paper as a = reject. Please address the questions listed in the weakness section *** Scientific/Technical Quality: How would you rate the = scientific/technical quality of the paper? (1 being worst score, 5 being = best score) Excellent (4) *** Innovation: How would you rate the innovation of the paper, the news = being received by the reader? (1 being worst score, 5 being best score) Average (3) *** Presentation Quality: How would you rate the quality of presentation = within the paper? (1 being worst score, 5 being best score) Excellent (4) *** Overall Recommendation: What is your over all recommendation for this paper? Accept (4)