----------------------- REVIEW 1 ---------------------
PAPER: 8
TITLE: On the Energy Efficiency and Performance of Neighbor Discovery Schemes for Low Duty Cycle IoT Devices
AUTHORS: Junaid Ahmed Khan, Romain Pujol, Razvan Stanica and Fabrice Valois

Overall evaluation: 2 (accept)

----------- Overall evaluation -----------
The paper proposes a comparison of neighbour discovery schemes with regards to energy efficiency and latency. The comparison is made theoretically, by simulation and by experimentation. 
The authors introduce two metrics to ease the comparisons and make it fair.

The results are interesting and kind of surprising since they show that some a priori ideas should be revised. 
I really appreciated the three aspects (theory, simulation and experimentation), which is a great strength of the paper. However, I would recommend authors to strengthen the comparison between them (there is no comparison with theoretical results. Comparison between simulation and experimentation is very light)

I have some minor comments and remarks to improve the paper. 
Fig 1a and 3a should represent the percentage of neighbour discovered to be more easily interpreted and for a better comparison. 
How to explain the irregularity for blinddate for theta ?
You talk about uneven discovery among nodes but this is not really shown. 

2nd paragraph of 2nd column of 2nd page does not end after the good sentence (?)

please, better motivate and explain why theta and delta are like this


our study also show --> shows


----------------------- REVIEW 2 ---------------------
PAPER: 8
TITLE: On the Energy Efficiency and Performance of Neighbor Discovery Schemes for Low Duty Cycle IoT Devices
AUTHORS: Junaid Ahmed Khan, Romain Pujol, Razvan Stanica and Fabrice Valois

Overall evaluation: 2 (accept)

----------- Overall evaluation -----------
Pe-WASUN_2017_paper_8

In this paper, the authors focus on neighbor discovery, in low duty cycle IoT systems. Moreover, they investigate the relation between energy efficiency, discovery latency and the fraction of discovered neighbors. 

The paper is well written and presented, while the studied research subject is very interesting.

My main concern for this work is that it remains to abstract regarding the employed protocols and standards, more specifically::

- It is not clear to me why the authors employ CSMA/CA if their focus is Low-power and low latency.

- The authors did not present the employed IoT stack. More specifically, they are working with 6TiSCH stack or some other Ad Hoc stack? 

- Then, which is the employ standard, i.e., IEEE 802.11 or IEEE 802.15.4-2015 TSCH?

- Next, why the authors calculate slots when they focus later in CSMA/CA approach where the time suppose to be continues and not slotted. 

Then regarding the performance evaluation (simulation), I am missing some information about the topology of the nodes, they are mobile or all static? If mobile then what is the speed of these nodes?