2011 International Conference on Advanced Technologies for Communications (ATC 2011)

Vietnam Spectrum Occupancy Measurements and Analysis for Cognitive Radio Applications Vo Nguyen Quoc Bao∗ , Le Quoc Cuong∗ , Le Quang Phu∗ , Tran Dinh Thuan∗ , Nguyen Thien Quy† , Lam Minh Trung† ∗ School

of Telecommunications Posts and Telecommunications Institute of Technology Email: {baovnq, lequoccuong, phulq, tdthuan}@ptithcm.edu.vn † Radio Frequency Directorate Vietnam Ministry of Information and Communications Email: {thienquy, trunglm}@rfd.gov.vn Abstract—The rapid growth of demand for wireless transmission has placed great pressure on the scarce radio spectrum due to the fixed spectrum allocation policy and cognitive radio (CR) is considered as a promising solution to such the problem. The main idea behind a cognitive radio network is for the unlicensed users (also called secondary users) to exploit opportunistically the underutilized spectrum licensed to a licensed network (referred to as a primary network). Before investigating the technical and political implications of CR, it is important to know to what extent the licensed bands are temporally unoccupied. In this paper, for the first time, we investigate the spectrum usage pattern in Vietnam in the frequency bands ranging from 20 MHz to 3000 MHz in Ho Chi Minh City and Long An province. The purpose of the measurement is to find how the scarce radio spectrum allocated to different services is utilized in Vietnam and identify the bands that could be accessed for future opportunistic use due to their low or no active utilization. These analyses indicate that, on average, the actual spectral usage in all bands is 13.74% and 11.19% for Ho Chi Minh City and Long An province, respectively. The experiment results also show that the spectrum band assigned for analog television (470-806) is the highest occupancy band with 58%.

I. I NTRODUCTION Radio frequency spectrum is a resource of fundamental importance in wireless communication systems. During recent years, a multitude of wireless applications and services has been developed resulting in a collision with the comprehensive, well-established, but increasingly obsolescent policy for the allocation and utilization of the radio spectrum [1]. In principle, international regulatory bodies like the Federal Communications Commission (FCC) coordinate and properly regulate the usage of radio spectrum resources and the regulation of radio emissions [2]–[4]. In Vietnam, based on the recommendations of the FCC, the radio frequency directorate (RFD) assigns spectrum to licensed holders, also known as primary users, on a long-term basis [5]1 2 3 . Such the policy partitions the overall radio spectrum into non-overlapping frequency bands corresponding to different purposes, i.e. government, public safety, national defense,

television radio, cellular, and unlicensed consumers. Although the fixed allocation approach ensures that competing wireless applications do not interfere with each other’s, it seems to be inappropriate since the current spectrum use is inefficient with some bands heavily subscribed and others rarely used. One of the most promising solutions for such the problem is cognitive radio, invented by Mitola [6], being able to detect an unoccupied frequency band to use temporarily, and then vacate when necessary [7]. To realize CR networks as well as to quantify their potential benefits accurately, the current spectrum usage should be known. Therefore, the crucial step is the study of the availability of temporarily unoccupied spectrum not only in terms of frequency but also in terms time and space. Till now, several measurement campaigns were conducted in USA [8], New Zealand [9], Germany [10], Singapore [11], China [12], Spain [13], and Qatar [14]. A shared finding among these studies is that a large portion of the assigned spectrum remains underutilized. In this paper, for the first time, we report the detail results of a spectrum survey conducted in Vietnam with an aim not only to fill the gaps today in knowledge about the use of spectrum but also to identify the most suitable and interesting bands for CRNs. The spectrum measurement campaign covers the frequency range from 20 MHz to 3000 MHz in two locations: Ho Chi Minh City and Long An province. The rest of this paper is organized as follows. In section II, we introduce the measurement setup and procedure. Section III presents the measurement results and provides some observation and discussion. Finally, the paper is closed in section IV. II. M EASUREMENT S ETUP AND P ROCEDURE The equipment used for the measurement consists of a set of antenna, an R&S EM550 VHF / UHF digital wideband receiver4 , and a server installed R&SARGUS monitoring software5 . The set of antennas including HE0166 , HE3097 , 4 http://www2.rohde-schwarz.com/product/em550.html

1 http://rfd.gov.vn/Danh

sach/Quy hoach moi 2 http://rfd.gov.vn/Danh+sach+Quy+hoach+bang+tan/Quy hoach bang tan 3 http://rfd.gov.vn/Danh sach/Quy

978-1-4577-1207-4/11/$26.00 ©2011 IEEE

135

5 http://www2.rohde-schwarz.com/product/ARGUS.html 6 http://www2.rohde-schwarz.com/product/HE016.html 7 http://www2.rohde-schwarz.com/product/HE309.html

HE015

HE309

HE314A1

HF214

HF902

Matrix Switch

Fig. 1.

The measuarement system.

Fig. 2.

HE314A18 , HF2149 and HF90210 is connected to the EM550 receiver via a switch matrix as shown in Fig. 1. Via the R&SARGUS monitoring software, we can configure the system to measure the received signal power in all range of frequency (of interest) from 20 MHz to 3000 MHz as well as stored all the measurement results in the server in real time for further processing. The detail characteristics of each antenna are provided in Table I. Depending on the measured spectrum band, the appropriate antenna is manually chosen, i.e. it is directly connected to the receiver. For example, we use an active vertical dipole (HE309) for the frequency range between 20MHz to 1300 MHz while HF902 should be used if the frequency range from 1 to 3 GHz is considered. The measurements were conducted over a 4-month time span from Oct. 2010 to Feb. 2011 at two different locations. The first location as shown in Figs. 2 is the roof top (the 6th floor) of the RFD office building in An Phu, District 2, Ho Chi Minh City with coordinate 10◦ 47 42.3 and 106◦ 44 25.9 . As a reference, the second location (Long An province with coordinate 10◦ 38 12.50 and 106◦ 29 36.00 ) is chosen roughly 50km far away the first location considered as a rural area. Both the chosen measurement sites almost have almost no high buildings surrounding enabling us to accurately measure the spectral activity of all possible transmitters. The utilization of spectrum is usually quantitatively examined by the most used metric - spectrum occupancy. In a particular location, it is defined as the probability that a measured signal of a certain bandwidth is unused by primary users. To determine whether a given frequency is occupied or not, spectrum sensing is usually used. There are three common methods proposed in the literature so far including matched filter detection, cyclostationary feature detection and energy detection [15], [16]. Among them, energy detection is the most popular method measuring only the received signal 8 http://www2.rohde-schwarz.com/product/HE314A1.html 9 http://www2.rohde-schwarz.com/en/products/antennas/HF214.html 10 http://www2.rohde-schwarz.com/en/products/antennas/HF902.html

The measurement system.

Fig. 3.

The antenna system.

power [17]. In this paper, we use the energy detection for all measurements. In energy detection approach, a frequency band is considered ”unused” if the power of measured signals is above a predefined power threshold, i.e. the noise level at the receiver. Otherwise, it is reported as an ”available” band for cognitive applications. Therefore, determining the level of background noise is a critical step towards spectrum occupancy measurement. According to [2], [10], [14], a margin of 3dB is considered for determining the final threshold value in order to account any unforeseen effects and variations. III. M EASUREMENT R ESULTS AND A NALYSIS In this section, we present some selective spectral measurement results. We start with the Fig. 5 and 6 where the average PSD over the whole frequency range of measurement study are plotted. Due to the characteristics of the antennae used, the measured frequency is divided into two non-overlapping

136

Fig. 4. The aerial map showing Ho Chi Minh City measurement site (Courtesy of Google Inc.). Type of antenna HE016

HE309 HE314A1 HF214 HF902

Antenna characteristics Active antenna system, omnidirectional reception of vertically and horizontally polarized signals with bandwidth 10 kHz to 80 MHz (vertical) and 600 kHz to 40 MHz (horizontal). Active vertical dipole, high sensitivity, large bandwidth and wide dynamic range from 20MHz to 1300 MHz. Active omnidirectional antenna, reception of horizontally polarized waves from 20 MHz to 500 MHz. Omnidirectional antenna, designed for the reception of horizontally polarized waves 500 MHz to 1300 MHz. Omnidirectional antenna designed for the reception of vertically and horizontally polarized waves from 1-3 GHz. TABLE I A NTENNA CHARACTERISTICS .

0

PSD [dBm]

−20 −40 −60 −80 −100 100

200

300

400

500 600 Frequency [MHz]

700

800

900

1000

Fig. 5. Average power spectral density vs. frequency at Ho Chi Minh City 20-1000 MHz. 0

PSD [dBm]

−20

−40

−60

−80

−100 1000

1200

1400

1600

1800

2000 2200 Frequency [MHz]

2400

2600

2800

3000

Fig. 6. Average power spectral density vs. frequency at Ho Chi Minh City 1000-3000 MHz.

frequency ranges, i.e. from 20 to 1000 Mhz and from 1001 MHz to 3000 MHz. From the figures, there are some important points that is worth noting as follows. The level of background noise is a little higher than the theoretical ambient noise. Importantly, it is not constant and slightly increases with frequency resulting in an increase on the decision threshold. It is due to the fact that the decision threshold is chosen 3-dB above the system noise floor as mentioned previously. Besides, we can also observe that the actual spectrum usage pattern is not uniform, i.e, the spectrum below 1 GHz seems to be heavily utilized while the spectrum from 2 to 3 GHz is found to be lightly used. Although Fig. 5 and 6 provide us some overview about spectrum utilization, they is insufficient to provide a detail description on how spectrum is used in different bands. Therefore, for a better view, we participate the overall spectrum under consideration into 15 sub-bands and study its associated spectrum activity in Figs. 7-22. In particular, Fig. 7, 8 and 9 show the received power versus frequency plot for 30-54 MHz, 54-68 MHz and 68-87 MHz, respectively. Clearly observed from the figures, the band allocated to FM radio is found to be most quiet as compared to the other bands. In Fig. 10, 12 and 15, we show the PSD in the broadcasting (FM, TV) bands: 87-108 MHz, 174-230 MHz and 490-806 MHz. As can be seen, they are the most heavily utilized bands observed in this study. The typical maximum signal power of FM bands is from 0 dBM to -20 dBm. With TV channels, the maximum power is around -60dBm to -40 dBm. From the figures, we can also identify some spectrum of signals coming from some popular FM and TV channels, e.g. VOV1, VOV2, VOV3, VOV5, FM99.9 , HTV7, HTV9, TayNinh, DongNai, etc. Along with the broadcasting bands, cellular mobile service bands (Fig. 14, 16 and 21) are the other ones having a considerably higher occupancy rate compared with other type of frequency allocations. If we consider land mobile bands 824-960 MHz and 1710-2300 MHz, it is easy to identify the spectrum of downlink GSM/E-GSM signals that are located in 950MHz and 1800MHz bands. Furthermore, the spectrum of downlink 3G/IMT2000 signals of four 3G service providers, i.e. Mobile, Viettel, EVN&HT, and Vinaphone, are observed ranging from 2110Mhz to 2200 MHz. It appears that the downlink channels in point-to-multipoint mobile applications are identified as mostly occupied. This is due to the active control channels constantly broadcasted by base stations to maintain cellular service coverage of GSM900, GSM1800 and WCDMA networks. Unlike downlink channels always transmitting with relatively high power, the usages in the uplink channels depend on the actual number of active mobile users in the measurement area and more intermittent according to their behaviors. From the figure, we can see that as expected transmit power of GSM900 mobile stations is higher than that of GSM1800 mobile stations. Besides, from Fig. 21, we also observe that 3G uplink channels seem to be completely unused. Here, it should be noted that due to the nature of WCDMA technology the transmit power of uplink channels

137

in 3G system is very low, and might not be detectable by the measurement system. Next, we focus on Fig. 11 and 13 where the spectrum activity of the frequency bands 108-174 and 230-406 is shows. As can be seen, most part of these bands band is unoccupied suggesting some opportunities for cognitive radio accesses. However, recalling that the whole band from 230 to 406 MHz is exclusively reserved for security services and systems of the Vietnam ministry of public security (MPS). Therefore, in principle, such spectrum bands should be precluded by secondary access. The average PSD of the well-known unlicensed Industrial, Scientific and Medical (ISM) band (2400 to 2500 MHz) is illustrated in Fig. 22. Allocated for unlicensed spectrum use, the ISM band is considered as the most open band. Within this band, many wireless applications are operated including WiFi transmitters, cordless telephones, microwave ovens, and various consumer products. Therefore, it is believed to be the most heavily used frequency band. However, from Fig. 22, it is observed that this band appears to be unoccupied. It can be explained by the fact that this frequency band is usually occupied in indoor environments and signals at such frequencies are severely attenuated by walls. The rest of spectrum between and 3 GHz remains mostly unused, with the exception of some signals with very low duty cycle in bands allocated to aeronautical and satellite radiolocation and radionavigation, (960-1350 and 1610-1710 MHz), DECT cordless phones (1880-1900 MHz) and military radars (27002900 MHz). To this end, we show the band-by-band average spectrum occupancy in Ho Chi Minh City in Fig. 23. The results of these measurements indicate that some spectrum bands are subjected to exhaustive usage while some others are sparsely used or show temperate utilization, and, in some cases, are not used at all. In general, the average spectrum occupancy observed in Ho Chi Minh City is 13.74% for the whole frequency range between 20MHz and 3000 MHz and the band assigned for television broadcasting is the highest occupancy band with 58%. Stated another way, 86.25% of this spectrum is unused. As a baseline for comparison, we also show the average spectrum occupancy of New York and Long An in Fig. 24. The obtained results demonstrate that Ho Chi Minh City spectrum utilization exceeds Long An by roughly 1.46%, which, in turns, exceeds New Yorks by 1.15%. IV. C ONCLUSION The spectrum measurements presented here are part of a larger on-going measurement campaign conducted by PTIT in several cities in the south of Vietnam. The purpose of this project is to create a usage map for cognitive applications in Vietnam. The challenge of this campaign is not only cost (equipment) but also time (deployment) where multiple locations are to be measured to obtain local spectral pattern usage. Our measurement results suggest that in Vietnam most of allocated frequencies are underutilized except for mobile and

broadcasting bands and CR applications can be realized by exploiting bands with low measured occupancy rates. However, care must be taken to account for possible wireless channel effects such as multi-path and hidden terminal problems. ACKNOWLEDGMENT This research was supported in part by the Science and Technology Foundation of Ho Chi Minh City and by the Vietnam’s National Foundation for Science and Technology Development (NAFOSTED) (No. 102.99-2010.10). R EFERENCES [1] Computer Science and Telecommunications Board, Wireless Technology Prospects and Policy Options. National Academy of Sciences, 2011. [Online]. Available: http://www.nap.edu/catalog.php?record id=13051# toc [2] R. Bureau, “Handbook on spectrum monitoring,” International Telecommunication Union (ITU), p. 168, 2002. [3] OFCOM, “United kingdom frequency allocation table,” 2010. [Online]. Available: http://tinyurl.com/44j8vpf [4] NTIA, “United states frequency allocations: The radio spectrum,” 2010. [Online]. Available: http://www.ntia.doc.gov/osmhome/allochrt.pdf [5] L. V. Tuan, “Radio frequency management in vietnam,” Journal on Information & Communications Technologies, 2008. [Online]. Available: http://www.tapchibcvt.gov.vn/vi-vn/sanphamdichvu/2007/8/18328.bcvt [6] I. Mitola, J. and J. Maguire, G. Q., “Cognitive radio: making software radios more personal,” IEEE Personal Communications, vol. 6, no. 4, pp. 13–18, 1999. [7] J. Wang, M. Ghosh, and K. Challapali, “Emerging cognitive radio applications: A survey,” IEEE Communications Magazine, vol. 49, no. 3, pp. 74–81, 2011. [8] M. A. McHenry, “Nsf spectrum occupancy measurements project summary,” Shared Spectrum Co., Tech. Rep., Aug. 2005. [9] R. I. C. Chiang, G. B. Rowe, and K. W. Sowerby, “A quantitative analysis of spectral occupancy measurements for cognitive radio,” in Proc. IEEE VTC2007-Spring Vehicular Technology Conf. (IEEE VTC 65th, pp. 3016–3020. [10] M. Wellens, J. Wu, and P. Mahonen, “Evaluation of spectrum occupancy in indoor and outdoor scenario in the context of cognitive radio,” in Proc. 2nd Int. Conf. Cognitive Radio Oriented Wireless Networks and Communications CrownCom 2007, pp. 420–427. [11] M. H. Islam, C. L. Koh, S. W. Oh, X. Qing, Y. Y. Lai, C. Wang, Y.-C. Liang, B. E. Toh, F. Chin, G. L. Tan, and W. Toh, “Spectrum survey in singapore: Occupancy measurements and analyses,” May 2008. [12] D. Chen, S. Yin, Q. Zhang, M. Liu, and S. Li, “Mining spectrum usage data: a large-scale spectrum measurement study.” ACM, pp. 13–24. [13] M. Lpez-Bentez, A. Umbert, and F. Casadevall, “Evaluation of spectrum occupancy in spain for cognitive radio applications.” IEEE, pp. 1–5. [14] K. A. Qaraqe, H. Celebi, A. Gorcin, A. El-Saigh, H. Arslan, and M.s. Alouini, “Empirical results for wideband multidimensional spectrum usage,” in Proc. IEEE 20th Int Personal, Indoor and Mobile Radio Communications Symp., pp. 1262–1266. [15] A. Goldsmith, S. A. Jafar, I. Maric, and S. Srinivasa, “Breaking spectrum gridlock with cognitive radios: An information theoretic perspective,” Proceedings of the IEEE, vol. 97, no. 5, pp. 894–914, 2009. [16] A. Ghasemi and E. S. Sousa, “Spectrum sensing in cognitive radio networks: requirements, challenges and design trade-offs [cognitive radio communications],” IEEE Communications Magazine, vol. 46, no. 4, pp. 32–39, 2008. [17] F. F. Digham, M. S. Alouini, and M. K. Simon, “On the energy detection of unknown signals over fading channels,” IEEE Journal on Communications, vol. 55, no. 1, pp. 21–24, 2007.

138

Amateur PSD [dBm]

0 Mean Min Max

−50

−100 30

35

40 Frequency [MHz] Fig. 7.

45

50

Amateur 30-54 MHz.

FM radio PSD [dBm]

0 Mean Max Min

−50 −100 54

56

58

60 F

Fig. 8.

62 [MH ]

64

66

68

Broadcasting (FM radio) 54-68 MHz.

Aero Nav., Analog TV. PSD [dBm]

0 Mean Max Min

−50

−100 68

70

72

74

Fig. 9.

76 78 Frequency [MHz]

80

82

84

86

Aeronautical Radionavigation/Analog TV 68-87 MHz.

FM radio PSD [dBm]

0 Mean Max Min

−50

−100 88

90

92

94

Fig. 10.

96 98 Frequency [MHz]

100

102

104

106

108

Broadcasting (FM radio) 87-108 MHz.

Aero, Mobile, Radio Nav., Maritime, Amateur, PMR PSD [dBm]

0 Mean Max Min

−50

−100 110

Fig. 11.

120

130

140 Frequency [MHz]

150

160

170

Aeronautical Radionavigation, Aeronautical Mobile (R), Mob. Sat. (S-E)/Fixed/Mobile, Amateur/Amateur Satellite, Fixed/Mobile 108-174 MHz.

139

Analog TV 0 PSD [dBm]

Mean

CH9:HTV9

CH7:HTV7

Min

CH11: TAYNINH

Max

CH12: DONGNAI

−50

−100 175

180

185

190

Fig. 12.

195

200 205 Frequency [MHz]

210

215

220

225

230

Broadcasting (TV channels 7 to 13) 174-230 MHz.

Fixed mobile, Aero, Others Data Transmisson of MPS

PSD [dBm]

0

−50

−100 240

260

280

Fig. 13.

PSD [dBm]

320 Frequency [MHz]

340

360

380

400

Fixed/Radionavigation Sat./Meteorological Aids 230-406 MHz.

Amatuer, fixed, mobile, radiolocation

DOWNLINK PMR TAXI

UPLINK PMR TAXI

0

300

CDMA 400/EVN

Mean Min Max

−50

−100 410

420

430

Fig. 14.

440 Frequency [MHz]

450

PSD [dBm]

CH21 CH22 CH23

470

Fixed/Radiolocation/Land Mobile 406,1-470 MHz.

Broadcasting 0

460

CH25 CH26

CH30

CH28

CH32

Mean CH34

CH36

Max

Min

CH39 CH40 CH41 CH42

−50 −100 480

PSD [dBm]

0

CH44

500

520 CH48

540

560 580 Frequency [MHz]

CH50 CH51 CH52

600

CH54 CH55 CH56

620 CH60

640 CH62

−50 −100 650

700

750 Frequency [MHz]

Fig. 15.

Broadcasting (TV channels 14 to 20)/Fixed 470-806 MHz.

140

800

DL DL DL DL E−GSM GSM GSM GSM VNMOB VINA VIETEL MOBI

Mobile Data, PMR & Trunking UL CDMACDMA S−PHONEE−GSM VNPT

UL GSM

−50

Mean Min Max

−100 820

840

860

880 900 Frequency [MHz]

Fig. 16.

920

PSD [dBm]

−55 DME 1003

940

960

Fixed/Land mobile 810-960 MHz.

Mean

DME DME 1024

DME 1201

Min

Max

DME 1260 DME 1280

DME 1340

−60 −65 −70 1000

1050

1100

Fig. 17.

1150 1200 Frequency [MHz]

1250

1300

1350

Aeronautical Radionavigation 960-1350 MHz.

Radio Location, Radio Astronomy, Earth Exploration Sat., Space Research PSD [dBm]

0 Mean Min Max

−20 −40 −60 1350

1360

1370

Fig. 18.

1380 1390 Frequency [MHz]

1400

1410

1420

Radio location/Earth Expl. Sat./Radio Astronomy 1350-1427.

Fixed −50 PSD [dBm]

PSD [dBm]

0

VIBA VNPT 1470.5 MHz

−55

Mean Min Max

−60 1430

1440

1450

1460

Fig. 19.

1470 1480 Frequency [MHz] Fixed 1427-1525 MHz.

141

1490

1500

1510

1520

Mobile Sat., Radio Astronomy, Meteorological Sat., Space Research, Fixed PSD [dBm]

−40

−60 1560

1580

1600 1620 1640 Frequency [MHz]

Fig. 20.

0 PSD [dBm]

Maritime Mobile Satellite (Earth to Space)

Maritime Mobile Satellite (Space to Earth)

−50

1540

1660

1700

Mean

GSM1800 (BT) Vina

Mobile Vietel GTel

1680

Mobile Satellite 1525-1660,5 MHz.

GSM1800 (BR) Vina

Max

Min

Mobile Vietel GTel

−20 −40 −60 −80

1750

1800

0 PSD [dBm]

Mean Min Max

Mobile Satellite

1850 Frequency [MHz]

1900

1950

2000

IMT2000

−20 −40 −60 2000

2050

Fig. 21.

2100

2150 Frequency [MHz]

2200

2250

2300

Radio Astronomy/Space Research/Meteological Aids/Mobile/Fixed 1710-2300 MHz.

ISM, Radio Location, Maritime Radionavigation, Aeronautical Radionavigation, Meteorological Aids. PSD [dBm]

0 Mean Min Max

−20 −40 2300

2400

Fig. 22.

2500

2600 2700 Frequency [MHz]

2800

Fixed/Mobile/Aeronatical navigation/Radiolocation 2300-3000 MHz.

142

2900

3000

Amateur 30−54 MHz Broadcasting (FM radio) 54−68 MHz Aeronautical Radionavigation/Analog TV 68−87 MHz Broadcasting (FM radio) 87−108 MHz Aeronautical Radionavigation 108−117,975 MHz Aeronautical Mobile (R) 117,975−137 MHz Mob. Sat. (S−E)/Fixed/Mobile 137−144 MHz Amateur/Amateur Satellite 144−146 MHz Fixed/Mobile 146−174 MHz Broadcasting (TV channels 7 to 13) 174−230 MHz Fixed/Radionavigation Sat./Meteorological Aids 230−406,1 MHz Fixed/Radiolocation/Land Mobile 406,1−470 MHz Broadcasting (TV channels 14 to 20)/Fixed 470−806 MHz Fixed/Land mobile 806−824 MHz Land Mobile/Fixed 824−960 MHz Aeronautical Radionavigation 960−1350 MHz Radio location/Earth Expl. Sat./Radio Astronomy 1350−1427 MHz Fixed 1427−1525 MHz Mobile Satellite 1525−1660,5 MHz Radio Astronomy/Space Research/Meteological Aids/Fixed 1660,5−1710 MHz GSM/Fixed 1710−2300MHz Fixed/Mobile/Aeronatical navigation/Radiolocation 2300−3100 MHz 0

Fig. 23.

30−54 MHz 54−88 MHz 108−138 MHz 138−174 MHz 174−216 MHz 216−225 MHz 225−406 MHz 406−470 MHz 470−512 MHz 512−608 MHz 608−698 MHz 698−806 MHz 806−902 MHz 902−928 MHz 928−906 MHz 960−1240 MHz 1240−1300 MHz 1300−1400 MHz 1400−1525 MHz 1525−1710 MHz 1710 1850 MHz 1850−1990 MHz 1990−2110 MHz 2110−2200 MHz 2200−2300 MHz 2300−2360 MHz 2360−2390 MHz 2390−2500 MHz 2500−2686 MHz 2686−2900 MHz 2900−3000 MHz

10

20

30

40

50

60

Average spectrum occupancy by band Ho Chi Minh City.

Ho Chi Minh City Long An New York 0

10

Fig. 24.

20

30

40

50

60

70

Average spectrum occupancy by band Ho Chi Minh City vs. Long An vs. New York.

143

80