Performance Evaluation of Directional Antennas in ZigBee Networks under NLOS Propagation Conditions
Round 1
Reviewer 1 Report
In the Abstract, define EIRP
PDR should be defined at the first place it has been mentioned to
Change EIRP (Effective Isotropic Radiated Power) to Effective Isotropic Radiated Power (EIRP)
Change CSMA 40 (Carrier-Sense Multiple Access) to Carrier-Sense Multiple Access (CSMA 40)
Change NLOS (Non-Line-Of-Sight) to Non-Line-Of-Sight (NLOS)
Change MAC 87 (Medium Access Control) to Medium Access Control (MAC87)
Change RSSI (Received Signal Strength Indicator) to Received Signal Strength Indicator (RSSI)
Change SINR (Signal-to-Interference-plus-Noise Ratio) to Signal-to-Interference-plus-Noise Ratio (SINR)
Change ISM (Industrial, Scientific, and Medical) to Industrial, Scientific, and Medical (ISM)
In Fig.*-b, Onidirectional should be corrected
Please, comment on the negative values of the SINR in Fig 8-b results. As it is well known, SINR values between 0 to 5 are considered poor.
In Fig.1-b, please put x-axis dimensions to show distance between transmitter and receivers.
In Fig. 1-b, you need to sketch the levels distance (vertical sketch) in meter, better than the description above the figure.
The Conclusion is not clear. The discussed results and conclusion contradict the expected results and aim in the Abstract.
Author Response
Thank you very much for your comments.
1) In the Abstract, define EIRP.
Answer: EIRP was defined in the abstract.
2) PDR should be defined at the first place it has been mentioned to.
Answer: PDR is first defined on line 106.
3) Change EIRP (Effective Isotropic Radiated Power) to Effective Isotropic Radiated Power (EIRP)
Change CSMA 40 (Carrier-Sense Multiple Access) to Carrier-Sense Multiple Access (CSMA 40)
Change NLOS (Non-Line-Of-Sight) to Non-Line-Of-Sight (NLOS)
Change MAC 87 (Medium Access Control) to Medium Access Control (MAC87)
Change RSSI (Received Signal Strength Indicator) to Received Signal Strength Indicator (RSSI)
Change SINR (Signal-to-Interference-plus-Noise Ratio) to Signal-to-Interference-plus-Noise Ratio (SINR)
Change ISM (Industrial, Scientific, and Medical) to Industrial, Scientific, and Medical (ISM)?
Answer: It was done as suggested by the reviewer.
4) In Fig.*-b, Onidirectional should be corrected
Answer: The word was corrected in Figure 8b.
5) Please, comment on the negative values of the SINR in Fig 8-b results. As it is well known, SINR values between 0 to 5 are considered poor.
Answer: According to figure 5b and figure 8b, the XBee receiver provided a 100% PDR for SINR around 0 dB. However, it is reaching the limit to start losing packets. Before understanding the results, it must be considered that the SINR values are estimates and not exact values because XBee modules do not measure SINR but only the RSSI.
ZigBee uses spread spectrum (DSSS) techniques, providing a processing gain of 9 dB over O-QPSK. From the datasheet, the minimum sensitivity of Xbee is -100 dBm for the PER of 1%. For packets of 128 bytes, it gives a bit error rate of about 10-5. The packet length used in the experiments was 40 bytes.
Also, as described in the paper “A Link Quality Inference Model for IEEE 802.15.4 Low-Rate WPANs“ (DOI: 10.1109/GLOCOM.2011.6133782), the ZigBee noise floor is -174-10*log10(BW)=-111 dBm for BW = 2 MHz. The minimum sensitivity is -111+NF+SINR=-100 dBm. For the EM357 radio, the NF (noise figure) is 9 dB, giving the minimum SNR of 2 dB. This means that packet losses occur for an SINR below 2 dB. We have SINR results around zero due to the smaller packet length and the use of error detection techniques in XBee modules through data retransmissions. Using the theoretical expressions of the error rate, a PER of 1% is obtained with an SINR = 1 dB for packets of128 bytes and no retransmission or with an SINR = -1 dB for packets of 128 bytes and 3 data retransmissions. For packets of 40 bytes, this PER is achieved with an SINR = -2 dB, in the line of the results presented in Figure 8b.
6) In Fig.1-b, please put x-axis dimensions to show distance between transmitter and receivers.
Answer: An x-axis was included.
7) In Fig. 1-b, you need to sketch the levels distance (vertical sketch) in meter, better than the description above the figure.
Answer: Figure 1 was changed to include the vertical dimensions. The description above the figure was also removed.
8) The Conclusion is not clear. The discussed results and conclusion contradict the expected results and aim in the Abstract.
Answer: After reading the Abstract, Discussion, and Conclusions, the authors do not find any contradiction. In the Abstract, we referred that the directional antenna does not improve the range over omnidirectional antenna under NLOS propagation conditions. The EIRP was the same for both antenna systems, as mentioned in the Abstract. In addition, control circuits may consume more power than omnidirectional antennas operating at higher power levels.
In Discussion and Conclusions, we evaluated the results obtained from this study and demonstrated the aims shown in the Abstract. Although the system with directional antennas has a gain of 7.5 dBi, this antenna could not improve the range over the omnidirectional one. The study of power consumption demonstrated the second assertion mentioned in the Abstract.
Reviewer 2 Report
The paper has an interesting topic. An acceptable experimental study has been done, which is useful in different industries and sections. However, the authors should address the following items to enrich the manuscript.
1. This is an interesting topic where the limitation of directional antennas are discussed in detail. Since directional antennas interference between networks operating in the same frequency channel. However after reading the overall paper it can be seen that this is a sort of survey study where directional and omni-directional antennas performance are studied. Authors need to be clarified that this is an original article or survey study. Since the proposed concept that omni directional antennas perform better is a well known concept.
2. In the introduction, as can be seen, few studies are related to the recent 3 years. Thereby, the authors are recommended to revise this section and add more recent studies to update the introduction.
3. In the introduction section, regarding the antenna and new technologies related to this area, the authors must use the following papers in order to enrich this section.
4-Port MIMO antenna with defected ground structure for 5G millimeter wave applications, Electronics 9 (1), 71, 2020.
Compact UWB band-notched antenna with integrated bluetooth for personal wireless communication and UWB applications, Electronics 8 (2), 158, 2019.
Integrated LTE and millimeter-wave 5G MIMO antenna system for 4G/5G wireless terminals, Sensors 20 (14), 3926, 2020.
Bandwidth enhancement and frequency scanning array antenna using novel UWB filter integration technique for OFDM UWB radar applications in wireless vital signs monitoring, Sensors 18 (9), 3155, 2018.
Resonator based switching technique between ultra wide band (UWB) and single/dual continuously tunable-notch behaviors in UWB radar for wireless vital signs monitoring, Sensors 18 (10), 3330, 2018.
A robust channel estimation scheme for 5G massive MIMO systems, Wireless Communications and Mobile Computing, 2019.
Malicious UAV detection using integrated audio and visual features for public safety applications, Sensors 20 (14), 3923, 2020.
A robust resource allocation scheme for device-to-device communications based on Q-learning, Computers, Materials & Continua 65 (2), 1487-1505, 2020.
4. The manuscript needs to be double-check in terms of english writing. There are many grammatical and other errors in the paper that must be revised.
5. Equation 3 has been taken from reference 27, which gives us an estimation for the I(w). How do the authors claim this? Why is the estimation of this equation better than the previous one? And how much is the error in this case?
6. In Figure 5, the omnidirectional has a weird behavior at the distance of 40 to 50m. It can be seen in both RSSI and PDR graphs. What is the interpretation of the authors for this behavior?
7. In the conclusion section, the authors have mentioned that "omnidirectional antennas spread more 457 power in space, and the 360 degree radiation pattern allowed for capturing more signals."
This is an obvious fact. Is this what the authors got from the study?
8. All in all, the paper does not recommend for the publication, and it needs major revisions.
Author Response
The paper has an interesting topic. An acceptable experimental study has been done, which is useful in different industries and sections. However, the authors should address the following items to enrich the manuscript
Answer: Thank you very much for your comments.
1) This is an interesting topic where the limitation of directional antennas are discussed in detail. Since directional antennas interference between networks operating in the same frequency channel. However after reading the overall paper it can be seen that this is a sort of survey study where directional and omni-directional antennas performance are studied. Authors need to be clarified that this is an original article or survey study. Since the proposed concept that omni directional antennas perform better is a well known concept.
Answer: This is an original article. The claim that omnidirectional antennas perform better requires a context. Omnidirectional antennas perform better if the objective is to transmit to many receivers simultaneously. Directional antennas perform better if the aim is to transmit to a single receiver. In wireless sensor networks, we have typically unicast transmissions. Omnidirectional antennas have been used for practical reasons, but directional antennas could perform better. The last one is the reason why many authors have proposed the use of directional antennas in WSN. Directional antennas allow increasing range and reduced interference. The problem is that this is valid for line-of-sight propagation conditions. To the best of our knowledge, no study with antennas was presented for non-line-of-sight propagation to demonstrate the conclusions of this work.
2) In the introduction, as can be seen, few studies are related to the recent 3 years. Thereby, the authors are recommended to revise this section and add more recent studies to update the introduction.
Answer: We have searched for actual studies related to the topic of this article. The presented list aims to show the main works in this area. We have included other recent articles in the references of the revised version.
3) In the introduction section, regarding the antenna and new technologies related to this area, the authors must use the following papers in order to enrich this section.
4-Port MIMO antenna with defected ground structure for 5G millimeter wave applications, Electronics 9 (1), 71, 2020. (Propose an antenna for 25.5–29.6 GHz frequency band for 5G applications. Not related with the present work.)
Compact UWB band-notched antenna with integrated bluetooth for personal wireless communication and UWB applications, Electronics 8 (2), 158, 2019. (Included)
Integrated LTE and millimeter-wave 5G MIMO antenna system for 4G/5G wireless terminals, Sensors 20 (14), 3926, 2020. (The proposed antennas were developed for LTE and 26–29.5 GHz frequency bands. Not related with the present work.)
Bandwidth enhancement and frequency scanning array antenna using novel UWB filter integration technique for OFDM UWB radar applications in wireless vital signs monitoring, Sensors 18 (9), 3155, 2018. (Propose an antenna for bandwidth enhancement and frequency scanning. Not related with the present work.)
Resonator based switching technique between ultra wide band (UWB) and single/dual continuously tunable-notch behaviors in UWB radar for wireless vital signs monitoring, Sensors 18 (10), 3330, 2018. (In the same line of the work included previously)
A robust channel estimation scheme for 5G massive MIMO systems, Wireless Communications and Mobile Computing, 2019. (Proposes a new channel estimation scheme applied a to MIMO systems, Not related with the present work)
Malicious UAV detection using integrated audio and visual features for public safety applications, Sensors 20 (14), 3923, 2020. (Deals with UAV detection. Not related to the work of the present study)
A robust resource allocation scheme for device-to-device communications based on Q-learning, Computers, Materials & Continua 65 (2), 1487-1505, 2020. (Proposed an algorithm to maximize throughput of device-to-device communications applied to cellular systems. Not related with the present work.)
Answer: The articles do not address the topics covered in this work. One paper was included in the section for new antennas. A comment was inserted in front of each proposed article.
4) The manuscript needs to be double-check in terms of english writing. There are many grammatical and other errors in the paper that must be revised.
Answer: The paper was revised, and the English was improved. A native speaker revised the manuscript.
5) Equation 3 has been taken from reference 27, which gives us an estimation for the I(w). How do the authors claim this? Why is the estimation of this equation better than the previous one? And how much is the error in this case?
Answer: The theory of the deconvolution technique to extract the received signal is presented in reference 27. In that work, it is explained that very high errors may occur if the extraction of I(w) is obtained from P(W)/G(W) because of the division by values close to zero of G(w) (we observed this problem). The filter used in that work to minimize the error committed in the extraction of I(w) was proposed by reference 28. In reference 27, the error was evaluated in the time domain for measurements performed in an anechoic chamber with some vegetation. The authors do not show information about the error results. They state that ideally, the average error should be zero, therefore, the absolute value of e needs to be minimized.
We used the frequency domain and the normalized error as reference 28. Converting equation (4) to the frequency domain, we may observe that the normalized error is only dependent on the antenna radiation pattern and the lambda parameter. In fact, the filter is used because the radiation pattern is not an ideal impulse. The error would be zero with an ideal impulse. This error is dependent on the lambda parameter. As explained in the paper, if lambda=0, the error is zero, but there is no filtering. If the lambda value is too high, the filtering effect is too high, distorting the results. The lambda parameter was obtained to produce a mean square of the normalized error of 10%.
6) In Figure 5, the omnidirectional has a weird behavior at the distance of 40 to 50 m. It can be seen in both RSSI and PDR graphs. What is the interpretation of the authors for this behavior?
Answer: In figure 5, the RSSI of the directional antenna decreases from a distance of 40 to 50 m and the PDR also reduces. For the omnidirectional antenna, the RSSI decreases, but the PDR practically maintains the value. The result seems strange but can happen in environments of NLOS or due to interference. The sensitivity of the Xbee is -100 dBm, but the minimum obtained values were -90 dBm due to Wi-Fi interference. Therefore, a higher RSSI may not mean more received power of the transmitted signal but more interference signal. Furthermore, the omnidirectional antenna receives more power from the space around the antenna, including directions from where the Wi-Fi is low.
Therefore, to understand the effect that occurred in the transmission from 40 to 50 m, we can use the results of the SINR shown in Figure 8b. These results were obtained by using the received signal from the generator. The results show that the SINR of the omnidirectional antenna system is better than for the directional one. As it is well-known, a small change in the SINR is enough to produce significant changes in the PDR near the sensitivity limit. The results confirm the omnidirectional antenna received less interference signal power than the directional one.
7) In the conclusion section, the authors have mentioned that "omnidirectional antennas spread more power in space, and the 360 degree radiation pattern allowed for capturing more signals."
This is an obvious fact. Is this what the authors got from the study?
Answer: Thank you for this comment. Reading this comment in the Conclusions seems to be a conclusion of the work. However, our intention was to conclude that because omnidirectional antennas spread more power in space, and the 360-degree radiation pattern allowed for capturing more signals, the received signal could reach the directional antenna's performance. In the revised version, we improved this conclusion sentence: “The results showed that the system with omnidirectional antennas performs better when receiving high signal components from various angles”.
Reviewer 3 Report
Performance evaluation of directional antennas in ZigBee networks under NLOS propagation conditions
In the study, a non-line-of-sight NLoS communication application measurements are discussed. In measurements, the ZigBee board is used. There are 2 antenna for comparison while one of them is omnidirectional, the other one is directional antenna. The study is interesting. The following suggestions will make the study more understandable and clear.
Reviews:
1- In the first use of an abbreviation, the expanded presentation of the abbreviation need to be presented first. The (equivalent isotropically radiated power) EIRP is used without any expanded presentation in line 18.
2- Between Line 37-41, there is a centence as follows: "The transmission required three parts. The first part was the preparation of the packet and the transmission to the radio chip buffer, which used 25% of the total time. The second part was the back-off timeout with duration dependent on the actual CSMA (Carrier-Sense Multiple Access) parameters, which used 56% of the total time. The last part was the transmission of the packet via the wireless radio, using 22% of the total time." The sum of these three part percentage is: 25% + 56% + 22% = 103% What does this mean?
3- Between Line 47 and 50: " The active period also consisted of three-time intervals. The activation and deactivation lasted 20 ms (8.1 mA), the listening and receiving ACK packets lasted 6 ms (40 mA), and the transmission of a 111-byte packet lasted 3.6 ms. The radios could op..."
4- During activation and deactivation period, duration and current consumption is presented. Similarly during listening and receiving ACK packets period same information presented but during transmission period the current consumption is not presented. Why the current consumption of this period is not provied?
5- In the paragraph where the received power measurement is mentioned, it is said that the antenna is attached to a stepper motor. Is the antenna that is attached to the stepper motor, receiver antenna or transmitter antenna. It is not clear from the paragraph. please clearly describe this situation.
6- In figure 5, are the points on graphic same with the points presented in figure 1? if so, please point the relation of them by labelling.
7- What are the used antennas? Besides the radiation pattern, are they dipole or array or reflector or yagi etc. They are only mentioned as directional and omnidirectional. What do you want to mean with orientation 1 and 2? Does it mean polarization? Please provide more information about them.
Author Response
In the study, a non-line-of-sight NLoS communication application measurements are discussed. In measurements, the ZigBee board is used. There are 2 antenna for comparison while one of them is omnidirectional, the other one is directional antenna. The study is interesting. The following suggestions will make the study more understandable and clear.
Answer: Thank you very much for your comments.
1) In the first use of an abbreviation, the expanded presentation of the abbreviation need to be presented first. The (equivalent isotropically radiated power) EIRP is used without any expanded presentation in line 18.
Answer: EIRP was defined in line 18.
2) Between Line 37-41, there is a sentence as follows: "The transmission required three parts. The first part was the preparation of the packet and the transmission to the radio chip buffer, which used 25% of the total time. The second part was the back-off timeout with duration dependent on the actual CSMA (Carrier-Sense Multiple Access) parameters, which used 56% of the total time. The last part was the transmission of the packet via the wireless radio, using 22% of the total time." The sum of these three part percentage is: 25% + 56% + 22% = 103% What does this mean?
Answer: Thank you for this correction. The results were corrected to 24%+54%+22%=100%. These results were obtained from a graph, and we cannot determine the exact values. However, the objective was to show the time distribution between transmission tasks.
3) Between Line 47 and 50: "The active period also consisted of three-time intervals. The activation and deactivation lasted 20 ms (8.1 mA), the listening and receiving ACK packets lasted 6 ms (40 mA), and the transmission of a 111-byte packet lasted 3.6 ms. The radios could op..."
4- During activation and deactivation period, duration and current consumption is presented. Similarly during listening and receiving ACK packets period same information presented but during transmission period the current consumption is not presented. Why the current consumption of this period is not provied?
Answer: The current consumption for the transmission part is dependent on the transmit power. However, as the reference authors presented a value for the transmit current, we also included the current consumption of this period in the revised version.
5) In the paragraph where the received power measurement is mentioned, it is said that the antenna is attached to a stepper motor. Is the antenna that is attached to the stepper motor, receiver antenna or transmitter antenna. It is not clear from the paragraph. please clearly describe this situation.
Answer: The sentence was completed to include the corresponding antenna: “We attached the receiving antenna to a stepper motor mounted on a pole to measure the signal distribution”.
6) In figure 5, are the points on graphic same with the points presented in figure 1? if so, please point the relation of them by labelling.
Answer: Following the suggestion of the reviewer, labelling was added to figure 5
7) What are the used antennas? Besides the radiation pattern, are they dipole or array or reflector or yagi etc. They are only mentioned as directional and omnidirectional. What do you want to mean with orientation 1 and 2? Does it mean polarization? Please provide more information about them.
Answer: The second paragraph of section 2.2 describes the used antennas. In line 168 it is referred that the omnidirectional antenna is a monopole of 2 dBi of gain. A reference to this antenna was included in the revised version. In line 170 it is described the directional antenna, consisting of a suspended patch. This antenna was built in our labs, and its dimension is presented in the paper. Following the suggestion of the reviewer, a new figure was included with the radiation pattern of this antenna (Figure 4b). The 17.5 dBi is an antenna array of 4x2 suspended antennas, whose radiation pattern is pretended in figure 4a.
Orientations 1 and 2 are described in the first paragraph of section 3.1. The first measurement described was to orient the transmit antenna towards the room door, approximately in line with most receiver positions (Directional - orientation 1). In the second case, we pointed the transmit antenna in a direction that provides the highest average RSSI and PDR (Directional - orientation 2). To make clear the reference to these situations, we changed the text to “Directional – orientation 1” and “Directional – orientation 2”. It was also included that the antennas operated in vertical polarisation.
Round 2
Reviewer 2 Report
The authors have answered most comments correctly.
Reviewer 3 Report
As a result of the re-review process, it was observed that all of the suggestions made were carefully considered. All changes made are clearly displayed and implemented correctly. In addition, the things that cannot be done are explained with their justifications and changes to correct the stated concerns are presented.
When the revised article is finally examined, it is deemed suitable for publication as is.
