Sub 1GHz Wireless LAN Deployment Scenarios and ... - IEEE Xplore

2 downloads 0 Views 611KB Size Report
In this study the design aspects of a new long-range outdoor Sub 1GHz ... such as Smart Grid, Surveillance, and Smart Farming and we will calculate link ...
IEEE Globecom 2011 Workshop on Rural Communications-Technologies, Applications, Strategies and Policies (RuralComm 2011)

Sub 1GHz Wireless LAN Deployment Scenarios and Design Implications in Rural Areas Stefan Aust1,Tetsuya Ito1 1

NEC Communication Systems, Ltd., 1753 Shimonumabe, Nakahara-ku, Kawasaki, Kanagawa 211-8666, Japan {aust.st, ito.tts}@ncos.nec.co.jp

Wireless local area networks (WLANs) can provide a very cost-efficient way of license-exempt wireless communication in outdoor regions with low population density. The use of carrier frequencies below 1GHz for wireless data transmissions would make such WLANs highly useful in rural areas due to an improved wireless propagation characteristic and larger coverage range compared to WLANs operating in 2.4GHz/5GHz band. The IEEE 802.11ah Task Group is going to specify a global WLAN standard that utilizes country specific carrier frequencies below 1GHz which are available for license-exempt wireless data transmission in various countries such as United States, Japan, China, Korea and Europe. In this study the design aspects of a new long-range outdoor Sub 1GHz WLAN system will be evaluated. We will discuss various use cases which can be realized in rural areas by using the Sub 1GHz band, such as Smart Grid, Surveillance, and Smart Farming and we will calculate link budgets and transmission outage probabilities to propose basic deployment guidelines for a license-free WLAN system that is designed to operate in the 900 MHz band. Index Terms—Sub 1GHz, WLAN, IEEE 802.11ah, coverage

I. INTRODUCTION

W

IRELESS local area networks (WLANs) and in particular the Wi-Fi certified WLAN systems have shown a tremendous success in the deployment in indoor and outdoor environments during the last decade. Indoor wireless networks have been the main Wi-Fi business market (indoor Wi-Fi, wireless office communication and campus networks). The deployment in large outdoor environments has shown a less demand due to the coverage of accessible cellular systems and WiMAX. The easy deployment, simple use, and high penetration of Wi-Fi interfaces in mobile communication devices increase the demand for outdoor deployment of ubiquitous wireless access. For instance, transmissions at very high data rate (e.g., up to 300Mbps using IEEE 802.11n) are possible at locations within close range of wireless access points. However, applications such as Smart Grid require a fast and simple deployment of long-range wireless communication networks for meter and sensor devices in rural areas. Current Wi-Fi systems are operating mainly at 2.4 GHz and 5GHz and are highly interfered by surrounding networks. Hence, an alternative frequency band will be required in the near future that is less interfered. Using ISM (Industrial, Scientific, and Medical) bands for license-exempt wireless communication below 1GHz could be a basis to design new WLAN systems, e.g., the IEEE 802.11ah WLAN standard. It is well understood that frequencies below 1GHz show good transmission characteristics compared to 2.4GHz systems due to the better propagation performance of lower frequencies in the UHF band (300MHz-3GHz) [1]. New ideas and use cases suggest using Sub 1GHz propagation performance for cost-efficient and large network deploying of Smart Utility Networks (SUN) at lower frequencies, e.g., in the TV White Space (TVWS) band [2]. For wireless systems based on IEEE 802.16e-2005

978-1-4673-0040-7/11/$26.00 ©2011 IEEE

(WiMAX) standard it has been found that at lower frequencies a significant better throughput coverage was observed at 825 MHz compared to higher WiMAX frequencies at 3535.5 MHz even with less power in areas with medium housing densities mainly due reduced scattering and attenuation [3]. Various terrain characteristics have been evaluated to design specific path loss models (e.g., line-of-sight, non-lineof-sight, non-flat) [4], [5] and diffraction loss optimized MIMO systems for rural and patchy wireless coverage environments, such as Australian outback [6], [7]. Several rural use cases have been proposed mainly for ehealth. In [8] the authors propose a cooperative Internet of Things (IoT) approach for a better health monitoring and in [9] the authors propose a low-cost mobile community e-health platform that operates in rural areas. In this study the aspects of a long-range outdoor WLAN system that will operate in rural areas will be studied. We discuss alternative use cases and the maximum size of wireless coverage for a wireless system that operates below1 GHz. We will calculate the link budget and we will consider system parameters such as frequency, path loss, bit rate and antenna height. Finally, outage characteristics will be discussed to estimate the number of associated stations to propose deployment guidelines for rural areas. II. SUB 1GHZ WIRELESS LAN SYSTEM Figure 1 shows an overview of wireless communication systems including the emerging WLAN standard IEEE 802.11ah [10]. The new WLAN system will operate at Sub 1GHz at country specific carrier frequencies. The relation between a low-cost wireless system and a large coverage make this system highly attractive for deployment in rural areas compared to cellular systems and WiMAX. IEEE 802.11ah is going to define an (MIMO) OFDM PHY that operates in the license-exempt bands below 1GHz in different countries.

1045

Fig. 1: Overview of wireless communication systems, including IEEE 802.11, cellular and WiMAX systems

Fig. 2: Smart Grid network in rural areas

Table 1 shows a list of countries and their related system frequencies [10]. The table also shows recent re-banding, e.g., in Japan, which harmonizes the spectrum at 920MHz globally. The IEEE 802.11ah MAC will provide mechanisms that enable co-existence with other systems including IEEE 802.15.4 and IEEE P802.15.4g [11]. IEEE 802.11ah aims to optimize the rate vs. range performance of the specific channelization in the country specific band. Main design features of IEEE 802.11ah are the support of an outdoor transmission range up to 1km and data rates at 100kbps or higher while maintaining the 802.11 WLAN user experience for fixed, outdoor, point to multipoint applications. Use cases include Smart Grid communications, cellular offloading and hot-spot coverage extension. The new WLAN standard can be easily integrated into current Wi-Fi chip designs [12] and performance enhancements related to speed and coverage by applying advanced antenna designs are being proposed [13]. III. USE CASES IN RURAL ENVIRONMENTS We present 3 use cases for rural Sub 1GHz communication systems in the following to identify the system requirements such as antenna height, bit rate, number of associated STAs and wireless coverage range.

A. Smart Grid Networks In the near future utility companies will start to equip their power line network infrastructure with additional sensors and meters. The goal is to make the entire utility grid greener by informing corporations and private end-users of their power usage, e.g., to reduce power consumption peaks which would lead otherwise to unstable operational network frequencies in the entire power line network [14]. In rural areas a wireless coverage range up to 1km is assumed, following the proposed wireless system in [10]. Sensors, such as power, gas, or water meter will require at least 100 kbps bit rate [14]. The proposed antenna height is between 5 and 15 meter. Meter-to-pole communication assumes that meter and sensors may be placed at the ground or even hidden behind meter boxes. Figure 2 shows a Smart Grid scenario consisting of power and gas meters with two orthogonal wireless networks. B. Surveillance Networks Surveillance networks play an important role in rural areas. The monitoring of obstacles by using cameras and motion sensors are necessary to sense changes of the safety for humans and value assets. Figure 3 shows a surveillance network that is applied in a rural area. Sensor-to-pole communication will be similar to Smart Grid, but a higher data rate such as 1000kbps is proposed, e.g., video or motion data.

Table 1: List of IEEE 802.11ah country specific carrier frequencies in Sub 1GHz band Country

Frequency [MHz]

Comment

United States

902-928

1W Tx power allowed

Japan

950-958

China

314-316, 430-434, 470-510, 779-787

Korea

917-923.5

Europe

863-870

Re-allocation to 915-930 MHz by 2012, 10mW Tx power allowed 470-562 MHz and 606-958 MHz covered by TV & broadcast band in China Only frequency bands >900 MHz are considered in [10] Only frequency bands>800 MHz are considered in [10]

Fig. 3: Wireless surveillance network in rural areas

1046

Fig. 4: Smart Farming network in rural areas C. Smart Farming Networks The deployment of sensors to detect temperature changes or outbreaks in livestock will become an important issue for farmers in the near future. Undetected flooding, fire or flu outbreaks can be devastating and entire farms can be destroyed within a short time. Figure 4 shows a Smart Farming scenario with a sensor deployment in an outdoor area to monitor temperature gradients, water levels and livestock conditions. Data rates are at 100kbps, AP elevation is at 5m, STA antenna elevation is at 0.5m-1m. Next we will discuss the required coverage and link budget for a wireless network deployment in rural areas. IV. DESIGN OF A SUB 1GHZ WIRELESS SYSTEM OPERATING IN RURAL AREAS A. Density, Coverage Radius, and Coverage Area The maximum number of wireless stations (STAs) which can connect to a WLAN access point (AP) depends on several factors, e.g., STAs density, signal power, path loss, and outage. First, we assume a simple omnidirectional radio propagation model as a basis to calculate the number of STAs based on the STA density and coverage radius. Second, we calculate the link budget to get an accurate number of STAs as a function of outage probability. Figure 5 shows an WLAN AP with 1 km coverage range and multiple STAs. The figure also indicates a STA density per square kilometer which is covered by the AP.

Fig 6: Number of STAs as a function of AP coverage range and STA density To calculate the number of STAs that is covered by the AP in Fig. 5 we apply a basic density calculation:

N sta

§ rcov erage · ¸¸ ˜ U sta © 1000 ¹

S ˜ ¨¨

(1)

with Nsta as the number of STAs within AP coverage, the AP coverage range rcoverage and STA density sta per square km. Fig. 6 shows the number of STAs within coverage vs. AP coverage range. The figure shows that for the assumed STA density for rural areas and 1km range almost 60 STAs are in coverage. The figure shows that if the coverage range reduces by 500m the maximum number of covered STAs in rural areas will be at 15 STAs. To compare the coverage range with other environments, the figure shows the results for sub-urban (500 STAs) and urban areas (2000 STAs) [14]. B. Calculation of Link Budget To understand the wireless link performance in a specific environment the calculation of the link budget or link margin should be calculated [1]. The link budget is a sum of related system parameters and includes the transmitter effective isotropically radiated power (EIRP), bit rate, the noise at the receiver and the path loss between transmitter and receiver. The path loss is of primary interest when calculating the link budget and includes the free-space loss, fading loss, and depends on the carrier frequency and environmental conditions. We use:

L path

GTx  GRx log10 d

(2)

with GTx as the transmitted signal power and GRx as the received signal as values based on the LTE macro formula in [15] and distance d between AT and STA in meter. The Link budget MB is (see table 2 for parameters):

MB Fig. 5: Single WLAN AP with 1km coverage range

1047

Ptx  J g  N 0  ha  Pi  Pm  10 log10 ( Dr ) (3)

Table 2: Applied system parameter and values for link budget calculation in (3) and (5) System parameter Carrier frequency fc [MHz] Antenna height ha [m] Tx power Ptx [dBm] Antenna gain Jg [dB] Shadowing Ps [dB] Mean noise N0 Bit rate Dr [kbps] Multipath fading loss Pm [dB] Implementation loss Pi [dB]

Applied values 900, 2400

3

Comment Link budget comparison at 900MHz/2.4GHz Network specific (picocell, macro-cell) Country/regulatory specific Regulatory specific

8

Environment specific

-170 100, 1000

Environment specific Application specific

3

Environment specific, line-of-sight Vendor specific

5, 15 0,10,20,30

3

Fig. 8: Maximum coverage range vs. Tx power (ha= 5m, Dr=100kbps)

Table 2 lists the selected parameter for the calculation of the link budget in (3) and the maximum coverage range in (5).

Figure 7 shows the result of the link budget calculation for two frequencies, at 900 MHz and 2400 MHz for comparing a Sub 1GHz system with a legacy Wi-Fi system operating at 2.4 GHz. The AP antenna height is set to 15m and the bit rate is assumed to be 100 kbps. The figure shows that the coverage almost doubles when using 900 MHz compared to 2.4 GHz. Especially for low sending power (1mW, 10mW) the effect of increasing the coverage about 100% shows an efficient way for designing a wireless communication system for rural areas. We can show that with 900 MHz a significant larger coverage area can be achieved. Next we show the result of the link budget calculation with antenna height set to 5 m. It can be observed in Fig. 8 that the maximum coverage range has been significantly reduced by 50% compared to Fig. 7. It is clear that the antenna height has a major impact on the coverage performance, and a high antenna elevation will lead to larger coverage areas. Next we discuss the link budget when the bit rate is set to 1000 kbps (AP antenna height is 15m). The higher bit rate is needed for the surveillance network use case.

Fig. 7: Maximum coverage range vs. Tx power (ha= 15m, Dr=100kbps)

Fig. 9: Maximum coverage range vs. Tx power (ha=15m, Dr=1000kbps)

An alternative way of calculating the performance of a wireless system is the link margin ML which is given by:

ML

EIRP  L path  G Rx  TH Rx

(4)

with EIRP in dBm, Lpath as the total path loss, including reflections and fade margins, GRx as the receive gain in dB, THrx as the receiver threshold in dB that will provide reliable operation. We will use ML for an outage calculation (a transmission is in outage, if the instantaneous SNR at the receiver falls below a specific SNR threshold). To calculate the maximum coverage range Cr we use:

Cr

10

§M ¨¨ ©

B

 G Tx  Ps G Rx

· ¸¸ ¹

(5)

1048

Table 3: Range and Number of associated STAs for Pout=0.001 and macro-cell deployment (ha=15m) Environment

Range [m]

Rural Sub-urban

Density [node/square km] 20 500

761 308

Number of STAs 36 162

Urban

2000

211

302

V. CONCLUSIONS AND FUTURE WORK

Fig. 10: Outage probability Pout for urban, sub-urban and rural areas (900 MHz) Figure 9 shows a dramatically reduced coverage range at 50% for increased bit rate compared to Fig. 7. It is clear that the bit rate has a significant impact on the coverage performance of a wireless system. The largest coverage area achievable by a specific technology requires transmission at the lowest bit rate used by that wireless technology.

The use of frequencies below 1GHz in future wireless systems is a promising way of deploying cost-efficient and long-range communication infrastructures in rural areas. We discussed the design of a long-range Sub 1GHz wireless system, in particular IEEE 802.11ah, and proposed 3 use cases to evaluate systems requirements such as the number of associated wireless stations, antenna height, and bit rate. Furthermore, we estimated the link budget and discussed deployment guidelines when a specific outage probability needs to be satisfied. In our future work we will evaluate further aspects of Sub 1GHz network planning including field tests of long-range WLAN systems in rural environments to verify the coverage range of Sub 1GHz wireless prototypes.

C. Outage Calculation In planning rural wireless communication systems a fundamental requirement is to obtain a specified minimum grade of service over the intended coverage area. Due to the fact that the signal is subject to multipath propagation, its amplitude at any given location can be expressed either as satisfactory or unsatisfactory reception, latter called as outage probability. We calculate the probability of the outage event Pout as the PDF of the received signal power s and apply the receiver threshold THRx from (4) as general definition of Pout as:

Pout : P>s  TH Rx @

TH Rx

³ ps s ds

REFERENCES [1] [2] [3]

[4]

[5]

[6]

(6)

0

[7]

In Fig. 10 the outage probability results are shown (antenna height=15m, bit rate=100kbps, tx power=20dBm), indicating a lower outage for rural areas due to the presence of dominant line-of-sight paths compared to sub-urban and urban areas. We set Pout = 0.001 and approximate the number of associated STAs and show the result in Table 3. The table shows that for rural environments with a density of 20 STAs the maximum coverage range is at 761m and a reduced number of STAs are in coverage, here only 36 STAs compared to Fig. 6. Similar reductions appear for sub-urban and urban areas by covering significant less STAs (162 STAs for sub-urban, 302 STAs for urban areas). As a guideline for a successful network planning we conclude that multiple APs need to be deployed, e.g., pico cells in urban areas, to reach a larger coverage in outdoor environments. We conclude that the results from Table 3 can be used as a guideline for the AP deployment in outdoor areas.

[8]

[9]

[10] [11] [12]

[13]

[14] [15]

1049

J. S. Seybold, “Introduction to RF Propagation”, Wiley, 2005. C.-S. Sum, H. Harada, et al., „Smart Utility Networks in TV White Space“, IEEE Communication Magazine, July 2011, pp. 132-139. C. Thein, A. Anggraini, et al., “Throughput coverage simulations based on signal level measurements at 825 MHz and 3535 MHz”, in 4th International Conference on Signal Processing and Communication Systems (ICPCS), December 2010, pp. 1-4. V. Dasarathan, M. Muthukuma, et al., “Outdoor Channel Measurement, Pathloss Modeling and System Simulation of 2.4 GHz WLAN IEEE 802.11g in Indian Rural Environments”, in Asia-Pacific Microwave Conference (APMC ‘07), Dec. 2007, pp. 1-4. M. Neunerdt, et al., “Land use classification as a key component for path loss prediction in rural areas”, in 7th International Symposium on Wireless Communication Systems (ISWCS), Sept. 2010, pp. 666-670. N. L. Ratnayake, et al., “Deterministic diffraction loss model for novel broadband communication in rural environments”, in IEEE Australian Communications Theory Workshop (AusCTW), Feb. 2011, pp. 49-54. J. Mazierska, et al., “Investigation into Improving Coverage and Capacity of CDMA Wireless Base Stations with CRFE in Rural Environment”, in Asia-Pacific Microwave Conference Proceedings, Dec. 2010, pp. 211- 214. V. M. Rohokale, et al., “A Cooperative Internet of Things (IoT) for Rural Healthcare Monitoring and Control”, in 2nd International Conference on Wireless Communication, Vehicular Technology, Information Theory & Electronics Systems, March 2011, pp. 1-6. Y. Su, et al., “Building an Information Quality Lab on E-health in the Rural Areas for Healthcare Education”, International Conference on E-health, Ecosystems and Technologies, April 2010, pp. 332-335. IEEE, “Sub 1GHz license-exempt PAR and 5C”, https://mentor.ieee.org/802.11/dcn/10/11-10-001-13. IEEE Std. 802.15.4, “Wireless MAC and PHY Specifications for Low-Rate WPANs”, 8 September, 2006. S. Datta, et al., “Pseudo Concurrent Quad-Band LNA Operating in 900 MHz/1.8GHz and 900MHz/2.4GHz Bands for Multi-standard Wireless Receiver”, VLSI Design, Jan. 2011, pp. 124-129. A. Budhawant, et al., “Wi-Fi Range and Speed Enhancement Using HHIS Based Steerable Square Loop Antenna”, in IEEE International Symposium on Antennas and Propagation Society, July 2010, pp. 1-4. NIST Priority Action Plan 2, Guidelines for Assessing Wireless Standards for Smart Grid Applications, ver. 1.0, December 31, 2010. V. Ponampalam, J. Wang, R. Porat, “TGah Outdoor Channel Models – Revised Text”, May 2011.