Towards Enabling Broadband for a Billon Plus Population

Towards Enabling Broadband for a Billon Plus Population
1
Towards Enabling Broadband for a Billon Plus
Population with TV White Spaces
arXiv:1603.01999v1 [cs.NI] 7 Mar 2016
Animesh Kumar, Abhay Karandikar, Gaurang Naik, Meghna Khaturia, Shubham Saha, Mahak Arora,
and Jaspreet Singh
Department of Electrical Engineering
Indian Institute of Technology Bombay
Mumbai 400076 India.
Contact Email: [email protected]
Abstract
One of the major impediments to providing broadband connectivity in semi-urban and rural India is the lack
of robust and affordable backhaul. Fiber connectivity in terms of backhaul that is being planned (or provided) by
the Government of India would reach only till rural offices (named Gram Panchayat) in the Indian rural areas. In
this exposition, we articulate how TV white space can address the challenge in providing broadband connectivity to
a billion plus population within India. The villages can form local Wi-Fi clusters. The problem of connecting the
Wi-Fi clusters to the optical fiber points can be addressed using a TV white space based backhaul (middle-mile)
network.
The amount of TV white space present in India is very large when compared with the developed world. Therefore,
we discuss a backhaul architecture for rural India, which utilizes TV white spaces. We also showcase results from
our TV white space testbed, which support the effectiveness of backhaul by using TV white spaces. Our testbed
provides a broadband access network to rural population in thirteen villages.The testbed is deployed over an area
of 25km2 , and extends seamless broadband connectivity from optical fiber locations or Internet gateways to remote
(difficult to connect) rural regions. We also discuss standards and TV white space regulations, which are pertinent
to the backhaul architecture mentioned above.
Index Terms
Wireless networks, wireless mesh networks, access networks, TV white spaces.
I. I NTRODUCTION
In the past decades, India has witnessed ever-increasing wireless telecom connectivity. There are over 940 million
wireless telecom subscribers with a tele-density of over 77, and India is the second largest telecom market in the
world. These numbers along with the advent of 3G and 4G systems would hint that wireless broadband access
in India has been solved; however, the reality is far from it! The number of broadband subscribers is around 86
million, and the rural tele-density is 46 (compared against the urban tele-density of 148). However, it can be stated
emphatically that the rural or affordable wireless broadband access is an unsolved problem in India.
Rural India has purchasing capacity since it contributes 50% to the GDP of India, though it has only 1.5%
registered broadband connections. It appears that rural broadband area is a largely untapped market with great
potential. However, there are significant challenges in providing broadband access in the rural areas, including: (i)
small average revenue per user as a fraction of total revenue; (ii) high capital and operation expenditure (including
license fees); (iii) affordable backhaul which is exacerbated due to a very large population, (iv) energy cost which is
worsened by lack of reliable power supply; and, (v) geographic accessibility issues such as right of way problems.
To alleviate the lack of broadband in rural areas, Government of India has been working with the initiative BharatNet
(formerly National Optical Fiber Plan or NOFN). Within BharatNet, which is being implemented in two phases,
point of presence (PoP) with optical connectivity at all village offices named Gram Panchayat will be provided. It
must be noted that mobility, at the moment, is not a major driver for broadband; instead, primary (fixed) broadband
service is the biggest requirement in rural India.
According to the National Telecom Policy, definition of broadband is a 2Mbps connection [1]. The current
broadband subscriber-base is around 15.35million in India. The targets of the Government of India are high and
2
mighty: by 2020, the Government of India plans to have a broadband subscriber-base of 600million. When coupled
with 2Mbps definition of broadband, a subscriber base of 250million in 2017, with 250GB/month at 2Mbps will
generate 100Exabytes of data per month in India alone. This is 8x larger than the expected global mobile traffic
by 2017! For example, consider the city of Mumbai, the population density is 21000/km2 . Approximately 34% is
wetland or forest. As a result, in some areas the population density is as high as 100000/km2 ! With an average
household size of 4, and if 100% homes have broadband, the amount of data generated will be 50Gbps/km2 .
Assuming 4 cells with radius < 500m, about 12Gbps per cell capacity will be required. The current wireless
technologies including Long Term Evolution (LTE)/LTE-A will be unable to address this. One of the solutions,
therefore, would be to deploy small cells such as dense Wi-Fi hotspots. However, due to non-availability of
ubiquitous fiber, backhauling of small cells is a challenge. We propose that TV UHF band radios can be used
to backhaul these dense Wi-Fi cells.
On the rural front, there are 250, 000 village offices named Gram Panchayat in India. The total number of
villages is 638, 619, so that each Gram Panchayat serves about 2.56 villages on an average. Each village has
around 4 hamlets at the periphery on an average. As mentioned earlier, BharatNet is an ambitious plan to provide
PoP at Gram Panchayat (offices) with optical fiber backhaul. Since the villages can be at a maximum distance of
a few kilometers from the Gram Panchayat, BharatNet will allay but not solve the problem of rural broadband in
India. In each hamlet or village, a wireless cluster can be formed (for example, by using a Wi-Fi access point),
but backhaul of the data from access points remains a challenge. We envisage that TV white spaces (in the UHF
band) can be utilized to backhaul data from village Wi-Fi clusters to the PoP provided by BharatNet.
This vision raises the following important questions—Can TV UHF band/TV white spaces be used to solve the
above-mentioned backhaul problem? How has the TV UHF band/TV white spaces utilized in the rest of the world?
How much TV white space is available in India and how does it compare with other countries? What network
topologies in the TV UHF band can be exploited to solve the backhaul problem? What results are obtained from
actual experimental testbed while performing backhaul in the TV UHF band over sparsely populated rural areas?
These questions will be subsequently answered in this article.1
II. TV WHITE SPACE OVERVIEW
With rising demand for bandwidth by various applications, researchers around the world have measured the
occupancy of spectrum in different countries. The observations suggest that except for the spectrum allocated to
services like cellular technologies, and the industrial, scientific and medical (ISM) bands, most of the allocated
spectrum is heavily underutilized (c.f. [3], [4], [5]). The overall usage of the analyzed spectrum ranges from 4.54%
in Singapore to 22.57% in Barcelona, Spain [3], [4]. Licensed but unutilized TV band spectrum is called as TV
white space in the literature [6], [7]. These white spaces in the TV UHF band (470-590MHz) have been of particular
interest owing to the superior propagation characteristics (from a received signal strength standpoint) [8]. Its status
in the world is reviewed next.
A. TV white space in various countries
The amount of available TV white space varies with location and time. The available TV white space depends
on regulations such as the protection margin given to the primary user, height above average terrain, transmission
power of secondary users, and separation of unlicensed user from the licensed ones. Since the actual availability of
TV white spaces varies both with location and time, operators of secondary services are interested in the amount of
available white space. TV white space estimation has been done in countries like the United States (US), the United
Kingdom (UK), Europe, Japan, and India [9], [10], [11], [12], [13]. For instance, in Japan, out of 40 channels, on
an average, 16.67 channels (41.67%) are available in 84.3% of the areas [12]. The available TV white space by
area in Germany, UK, Switzerland, Denmark on an average ranges between 48% to 63 out of the 40 TV channel
bands [11]. It must be noted that in these TV white space studies, the IMT-A (698-806MHz) band is also included.
FCC regulations in the US and Ofcom regulations in the UK have allowed for secondary operations in the TV
white spaces [6], [7]. For example, FCC regulations declare a band as unutilized if no licensed-user (primary) signal
1
In a mobility driven setup, TV white space testbed has been used in Microsoft Corporation, Redmond to backhaul data from a Wi-Fi
cluster in a moving shuttle [2].
3
is detected above a threshold of −114dBm [6]. Under this provision, a secondary user can use the unutilized TV
spectrum provided it does not cause harmful interference to the TV band users and it relinquishes the spectrum
when a primary user starts operation.
B. Standards and technologies to address TV white spaces
IEEE 802.11af has been designed by extending IEEE 802.11ac to support 8MHz channels and the use of a TV
white space database to inform and control the use of spectrum by the devices [14]. IEEE 802.11af uses carrier sense
multiple access at both the base station and the clients to share the spectrum. The spectral efficiency of 802.11af
in a single antenna configuration varies from 0.3bits/s/Hz to 4.5bits/s/Hz, resulting in a maximum throughput of
35.6Mbps over an 8MHz channel. IEEE 802.11-based technologies including 802.11af and its variants, operate at
a power level of 100mW-1W, and have a range of 1km in a last mile setting; and, it has been shown to work up
to 15km with 10dBi sectoral antennas in a middle-mile setting.
IEEE 802.22 is another IEEE standard designed for enabling broadband wireless access in TV white spaces [15].
IEEE 802.22 uses Orthogonal Frequency Division Multiple Access as the medium access (MAC) layer and uses
centralized scheduling of the MAC resources by the base station. The spectral efficiency of IEEE 802.22 in a single
antenna configuration varies from 0.6 bits/s/Hz to 3.1 bits/s/Hz, with a maximum throughput of about 19Mbps in
a single channel.
IEEE 802.15.4M is geared towards low-rate wireless personal area networks, with applications that include
machine to machine networks. A task group IEEE 802.15.4m is addressing which technologies should be enabled
in the TV white spaces. IEEE 802.19 is a standard for enabling co-existence between different technologies, with
specific focus on TV white spaces [16]. IEEE 802.19 defines an architecture and protocols for enabling co-existence
between different secondary networks. Finally, 1900.7 has been established for advanced spectrum management
and next generation radios.
In the future, we expect more technologies to be designed for operation over TV white spaces including LTE [17]
and IEEE 802.11ah. We note here that LTE technology can be adopted for TV white space by introducing design
changes to the RF transmitter (for example, by tuning the RF section to operate in TV UHF band).
C. Geolocation database and white space device access rules
To ensure coexistence of the TV broadcasters with the secondary devices, geolocation databases have been
mandated by the regulators FCC and Ofcom [6], [7]. All devices should have a location accuracy of ±50m and
must query a certified TV white space database to obtain an allowable channel with associated transmit power. The
list of available (unutilized by primary) channel, the channel access schedule for 48hours, and the transmit power
that is allowed is provided by the TV white space database.
D. TV white space availability in India
In India, the sole terrestrial TV service provider is Doordarshan which currently transmits in two channels in
most parts across India. Currently Doordarshan has 373 TV transmitters operating in the TV UHF band (more
precisely, TV UHF band-IV in 470-590MHz) in India. The TV UHF band consists of fifteen channels of 8MHz
each. In India, a small number of transmitters operate in this TV UHF bands; as a result, apart from 8-16MHz
band depending on the location, the UHF band is not utilized in India! Comprehensive quantitative assessment and
estimates for the TV white space in the 470-590MHz band for four zones of India have been presented by us in
the literature [13]. It has been shown that in almost all cases at least 12 out of the 15 channels (80%) are available
as TV white space in 100% of the areas in India [13] (see Fig. 1).
E. TV UHF band utilization in India and its policy aspects
TV white space in India is significantly larger than other countries reviewed above. The common approach for TV
white space utilization is through the use of whitespace devices and associated country-specific regulations [6], [7].
Such white space devices and regulations utilize the presence of database lookup, with transmit power limitations on
the unlicensed user. While 470-590MHz band has been licensed to TV broadcasting, its usage for rural broadband
4
Fig. 1.
TV white space available in India is illustrated. Observe that in almost all of the places, 100MHz of spectrum is nearly unutilized!
can be fundamentally different in India than through TV white space regulations. This fundamental difference is
explained next.
In the National Frequency Allocation Plan (NFAP) of 2011 [1], the spectrum in the frequency band 470-890MHz
is earmarked for Fixed, Mobile and Broadcasting Services. India belongs to Region 3 of the ITU terrestrial spectrum
allocations. In the 470-590MHz band, ITU permits Fixed, Mobile, Broadcasting services in the Region 3 [1]. As per
India footnote 20 [1], fixed services in the 470-590MHz band are allowed in India within the existing ITU spectrum
plan. This is in contrast with Region 1 (including Europe) where only broadcasting services are allowed in this
band, and Region 2 (including United States) where fixed services are allowed only in 470-512MHz. Accordingly,
fixed services can be allowed in 470-590MHz in India. This difference accommodates high-power transmissions
by any fixed service (such as broadband base stations) in the 470-590MHz band in India.
We suggest a license-exempt registered-shared-access based regulatory approach. Operators using TV white space
spectrum may register with a database and may have to share a channel or a sub-channel with other users on the same
channel in the vicinity. Further, different registered operators must cooperate and coexist to ensure high average
spectral efficiency when compared to random access. Techniques such as LBT and other suitable inter-operator
co-operation techniques can be combined with database assistance.
III. A
BROADBAND ACCESS - NETWORK TOPOLOGY FOR RURAL I NDIA
The 470-590MHz band, henceforth the UHF band for brevity, in India is heavily underutilized [13], and its radio
propagation characteristics are much better than and unlicensed band such as 2.4GHz [8]. In fact, its propagation
characteristics are suitable for non-line of sight connectivity. It is envisaged that a broadband access network can
be provided by extending Internet coverage from a rural PoP provided by BharatNet (an optical fiber point), by
using TV white space in the UHF band. In such a scenario, broadband base stations operating in the UHF band will
provide backhaul from villages to the PoP provided by BharatNet. Each village can be served by an unlicensed-band
Wi-Fi cluster. This architecture can be used to provide affordable broadband access-network in (rural) India, and it
results in a mesh-network of nodes operating in the UHF band as illustrated in Fig. 2. The typical distance between
nodes operating in 470-590MHz is around 1-5km.
In summary, the PoP locations are provided by BharatNet. The villages or hamlets can connect to wireless access
points using 2.4GHz Wi-Fi, which is an affordable short-distance technology. The end-devices (for economy of
scale and for ubiquity) will connect to the UHF band mesh network via the collocated 2.4GHz Wi-Fi access-points.
At each Wi-Fi access-point, a UHF band node will be provisioned. Then UHF band nodes can be used to backhaul
the data from these Wi-Fi access points to the PoP locations. The TV band base stations or relays can connect
in different topologies: (i) point to point; (ii) multipoint to point; and (iii) multi-hop mesh network. Of these, we
explore the most general topology of mesh-network.
5
Fig. 2. It is assumed that there is a PoP with a UHF band node, to provide broadband access-network in nearby villages. Geographically
sparse and distributed villages will form local 2.4GHz Wi-Fi clusters with Wi-Fi access points. Collocated with each Wi-Fi access point, a
UHF band node (a client or a base station) will be provisioned. The data from the Wi-Fi networks will be backhauled to the PoP by these
UHF band nodes.
The advantages of this broadband access-network are as follows: (i) the access technology Wi-Fi is affordable
for the rural population in India; (ii) the backhaul is cheaper due to TV UHF band propagation characteristics
(the license aspects can be handled using registered shared access, where multiple operators can coexist through
database lookup [18]); and, (iii) the power consumption of each UHF band node is low (5-10W in our testbed),
so that it can be powered through battery or solar energy. The next section highlights the results of our testbed
implemented in the multi-hop mesh topology, in a cluster of villages near Mumbai, India.
IV. T HE F IRST TV
WHITE SPACE
T EST-B ED IN I NDIA
An instance of the mesh-network proposed in the previous section has been realized in a testbed at Khamloli,
Maharashtra, India. The testbed is located in the Palghar district at about 107km distance from Indian Institute of
Technology Bombay. It is the first TV white space testbed in India. The layout of villages surrounding Khamloli is
illustrated in Fig. 3. The typical distance between two villages is in the range of 5km as mentioned, and the testbed
is deployed over an area of 25km2 . Khamloli village node forms the PoP in this layout with a 20Mbps optical fiber
link. With Khamloli village as PoP, a mesh-network is setup to extend Internet from Khamloli (wirelessly in the
UHF band) to the surrounding villages. The testbed provides a broadband access-network to the rural population
in thirteen villages or hamlets. Our testbed consists of 10 UHF band nodes functioning as client and 1 UHF band
node as base-station. The clients in UHF band are connected to Wi-Fi hotspots to provide Internet access.
A. Topographic details of the testbed
To give an idea about the propagation characteristics of RF signal, the topographical settings of various tower
locations and UHF band nodes are described. It must be noted that there are significant differences in the topographic
settings between Khamloli (PoP) and the other towers. There are four UHF band nodes setup as base-station nodes.
These result in Khamloli-Maswan, Khamloli-Haloli, Khamloli-Ganje, and Khamloli-Pargaon (see Fig. 4(a)) links.
There is no line of sight in Khamloli-Maswan and Khamloli-Ganje links. There is heavy vegetation in KhamloliGanje link. All other links have moderate vegetation.
There are seven UHF band nodes setup as client named Khamloli 1, Khamloli 2, Bahadoli 1, Bahadoli 2,
Dhuktan 1, Dhuktan 2, and Dhuktan 3. These UHF band nodes are setup at a height ranging in 4-6meters, with
small houses in between. Khamloli 1, Khamloli 2, Bahadoli 1, and Bahadoli 2 have line of sight with Khamloli PoP.
6
Fig. 3.
An example of Indian rural topology is illustrated. Khamloli village has a 20Mbps optical fiber link to the Internet. A wireless
mesh-network can be setup with Khamloli as the PoP to provide a broadband access-network in the nearby villages, while leveraging on the
unutilized UHF band.
But, Dhuktan 1 and Dhuktan 3 do not have line of sight while Dhuktan 2 has partial line of sight with Khamloli
PoP. Vegetation is large only in Dhuktan 3 link (see Fig. 4(b)).
Fig. 4.
Khamloli is a PoP in our testbed. (a) The long-distance links of the testbed are illustrated. Each tower is equipped with a UHF
band base-station. (b) The short-distance links of the testbed are illustrated. Khamloli tower has a UHF band base-station, while other nodes
have UHF band clients.
The UHF band node at Khamloli PoP is a base-station, has an omnidirectional antenna, and is mounted at a
height of 30 meters. At each location, a suitable site was identified for setting up the UHF band equipment and
associated accessories including power supply. At every location marked in the testbed layout (see Fig. 4(b)), Wi-Fi
routers have been installed to test the Internet connectivity. The customer premises equipments are Wi-Fi enabled
devices such as tablets or smartphones.
7
B. Our developed economical UHF band node prototype
Products based on IEEE 802.22 and IEEE 802.11af standards are available off-the shelf, but are expensive
(USD4000-5000 for base-station and USD1000-2000 for client). From an affordable broadband service point of
view, we decided to develop a low-cost prototype of UHF band node; and, our developed prototype costs USD650
per UHF band node. We adopted the approach of using standard IEEE 802.11g Wi-Fi with radio-frequency of
500MHz. Our device has OpenWrt operating system ported on it. This enabled us to implement the Protocol to
Access White Space (PAWS), an Internet Engineering Task Force (IETF) standardized protocol. 2 Our prototype
comprises of a commercially available off-the-shelf Wi-Fi routerboard that can connect with an RF card having a
mini PCI interface. The RF card converts the baseband signal to 2.4GHz, which is translated to 500MHz (UHF
band) by a downconverter. Few pictures from our testbed illustrating the box of prototype are shown in Fig. 5.
Fig. 5. Photographs of some UHF band nodes (clients) from Fig. 4 are shown. (a) Dhuktan 2 client is mounted at the top of a hill with
no houses around. (b) and (c) UHF band clients are typically mounted at a height of 4-6meters.
C. The Coexistence Handler–a UHF Band Database for India
The testbed’s UHF band devices query an OpenPAWS database setup by our research group to select the frequency
of operation, to avoid interference to any terrestrial TV services in other geographical regions of India [19].
OpenPAWS client was implemented in OpenWrt (ported on TV UHF band node), while the OpenPAWS server was
implemented in Linux. The OpenPAWS based database server implementation was tested to ensure its functioning
with our testbed. An error message is generated by the OpenPAWS server in the following three scenarios: (i) the
UHF band device is outside the regulatory domain of the TV white space database; (ii) the UHF band device sends
an AVAIL SPECTRUM REQ before initialization and registration; or, (iii) there are no channels available at the
location of the UHF band device. In our setup, the locations of the UHF band devices are known since it is a fixed
network. This rules out error in (i). The error in (ii) can be controlled by proper programming. In India, it has been
shown that there is a channel available at any point [13]. Therefore, OpenPAWS server’s response will be a list of
available channels with transmit power allocation.
For the Khamloli village (PoP) in the testbed, the OpenPAWS database declared Channel 1 to Channel 15 as
available since the Channels are available for transmission.3 The transmit power was restricted to 30dBm. After
assigning the power and channel to various UHF band nodes, the database is updated to reflect the same. The
2
We also experimented with IEEE 802.22 standard devices (such as those available from Carlson Wireless), however, large scale deployment
of testbed did not use them.
3
Note that any Channel in 470-590MHz could be used since the entire band is free in the vicinity of Khamloli!
8
database created is open for public access [19] and can be used to view the list of all TV towers operating in India
in the UHF band along with all operational parameters. It also displays TV towers operating on any particular
channel along with the tower’s coverage area calculated in [13].
D. Results obtained from the testbed
The results that were obtained with our testbed after extensive experimentation over a few months are highlighted
in this section. These comprise of throughput analysis for various links, and the variation and latencies observed
in the throughput. An ATM machine use-case is also explained towards the end of this section.
There are many point to point links in our testbed, from which we exemplify one. This is a long-distance with
line of sight (Khamloli-Pargaon). For a bandwidth of 5MHz, the obtained uplink and downlink throughputs (while
using UDP as well as TCP) are illustrated in Fig. 6. In the plots, observe that the received signal to noise ratio (SNR)
varies since the transmit power is kept constant (at 27dBm with 8dBi antenna gain) and the receiver’s distance and
associated conditions vary. Consistently a throughput of 4-5Mbps (in TCP) and 5-6Mbps (in UDP) were obtained
for the uplink as well as downlink. The tool used to monitor the network was iperf. TCP and UDP throughputs
over each link were computed using 500 different samples.
Fig. 6.
The elevation profile of Khamloli-Pargaon link, and the obtained throughput for TCP/UDP protocols are shown. The channel
bandwidth was 5MHz, and the transmit power and antenna gain were 27dBm and 8dBi, respectively.
The variation of latencies was also examined by 25000 randomly taken measurements. Two extreme-range of
wireless topographies were considered—Khamloli-Ganje which represents a 6.7km long-distance link at 5MHz
bandwidth (large distance and low bandwidth), and Khamloli-Dhuktan which represents a 2.3km moderate-distance
link at 20MHz bandwidth (small distance and large bandwidth). The latencies for Khamloli-Ganje link varies in
2-15ms, while its UDP throughput varies in 5.6-8Mbps. The latencies for Khamloli-Dhuktan link varies in 2-11ms,
while its UDP throughput varies in 11-17Mbps. These results are very exciting and promising!
High speed Internet access has been provided to the villagers in Khamloli, Bahadoli and Dhuktan using Wi-Fi
hotspots deployed at 10 locations across 3 villages and 3 kiosks one in each village. About 60 villagers from
three villages and surrounding hamlets (pada) have been trained as ‘e-Sevaks’ (electronic serviceman). An e-Sevak
assists other villagers in using Internet services for simple tasks like filling college forms, paying electricity bills,
and booking of train tickets. Kiosks set up in the three villages are run by these e-Sevaks on a daily basis for three
hours for this purpose. E-Sevaks have been given tablets for getting familiar with these Internet services.
An automated Teller Machine (ATM) provided by TATA Indicash is setup at Dhuktan Gram Panchayat (village
office) in order to demonstrate the use of e-Finance capabilities of TV white space for the rural areas. The ATM
machine is on a different private LAN which is tunneled through a Border Gateway Protocol (BGP) route from
Khamloli PoP to the ATM network. Tata Teleservices Limited and Tata Communications provisioned the MultiProtocol Label Switch (MPLS) leg of 64kbps at Khamloli, which was connected to security gateway application
from Pfsense (http://www.pfsense.com) by configuring one of the optional ports as a wireless area network
(WAN). The ATM machine connects to a Wi-Fi point to point link by Ethernet; this point to point link is connected
to a Wi-Fi access point, which connects with the PoP at Khamloli via TV white space network.
9
V. C ONCLUSIONS AND F UTURE W ORK
In this paper, we have articulated how UHF band can address the challenge in providing broadband connectivity
to a billion plus population of India. As outlined in the paper, one of the major impediments to providing broadband
connectivity in semi-urban and rural India is the lack of robust and affordable backhaul. Even in urban areas, one
of the major impediments for widespread deployment of Wi-Fi Hotspots is the lack of connectivity from Wi-Fi
access points to optical fiber gateways. Fiber connectivity in terms of backhaul that is being planned (or provided)
by the Government of India would reach only till the Gram Panchayat in the rural areas. In such a scenario, the
problem of connecting the Wi-Fi clusters to the optical fiber PoP can be addressed using a TV white space based
backhaul (middle-mile) network.
We believe that a cost effective solution for backhaul would require a license exempt database assisted approach
for TV white space spectrum management. Since UHF band is sparsely utilized in India by the broadcaster,
the challenge is not primary-secondary co-existence (as in many countries) but secondary-secondary co-existence.
Multiple operators should be able to share the TV spectrum and co-exist. While listen before talk (LBT) approach as
in IEEE 802.11af standard is one of the options for co-existence, it will pose challenges in rural regions with large
cell radius, due to superior propagation characteristics of sub-GHz spectrum. A combination of database assisted
and LBT is a topic for future investigation for providing “primary” broadband connectivity by possibly many local
operators.
Dynamic resource allocation algorithm with fair sharing of resources between multiple operators (co-primary)
is another area that requires more attention.Since TV band network is being proposed as middle mile network for
backhauling Wi-Fi clusters,both Wi-Fi access points and TV band radios can be controlled through a Software
Defined Networking (SDN) controller. This SDN controller can also be integrated with Database controller. Finally
SDN controller can also be employed to dynamically configure flow routing in a more complex topology of multihop mesh based middle mile. A SDN enabled Policy based Radios deployed for middle mile fixed services can set
the vision for 5G for India. Recently, the need for such vision for 5G has been also articulated by Eriksson and
van de Beek [20].
We are currently investigating these topics as enhancements to existing TV white space standards. These topics
are also under discussion in Telecom Standards Development Society of India (http://www.tsdsi.org).
R EFERENCES
[1] WPC, “National Frequency Allocation Plan 2011,” Department of Telecommunication, Government of India, Tech. Rep., 2011.
[2] R. Chandra, T. Moscibroda, P. Bahl, R. Murty, G. Nychis, and X. Wang, “A campus-wide testbed over the TV white spaces,” ACM
SIGMOBILE Mobile Computing and Communications Review, vol. 15, no. 3, pp. 2–9, 2011.
[3] 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,” in Proc. of the third Cognitive Radio Oriented Wireless Networks and
Communications, May 2008, pp. 1–7.
[4] M. López-Benı́tez, A. Umbert, and F. Casadevall, “Evaluation of spectrum occupancy in Spain for cognitive radio applications,” in
Proc. of IEEE Vehicular Technology Conference. IEEE, New York, Apr. 2009, pp. 1–5.
[5] R. I. C. Chiang, G. B. Rowe, and K. W. Sowerby, “A quantitative analysis of spectral occupancy measurements for cognitive radio,”
in Proc. of the sixty fifth IEEE Vehicular Technology Conference, April 2007, pp. 3016–3020.
[6] FCC 08-260, “Second report and order and memorandum opinion and order,” US Federal Communication Commission, Tech. Rep.,
Nov. 2008.
[7] Ofcom, “Digital dividend: Cognitive access. consultation on license-exempting cognitive devices using interleaved spectrum,” Ofcom,
UK, Tech. Rep., Feb. 2009. [Online]. Available: http://stakeholders.ofcom.org.uk/binaries/consultations/cognitive/state-ment/statement.
pdf
[8] J. D. Parsons, The Mobile Radio Propagation Channel Vol. 2. New York: Wiley, 2000.
[9] K. Harrison, S. M. Mishra, and A. Sahai, “How much white-space capacity is there?” in Proc. of IEEE New Frontiers in Dynamic
Spectrum. IEEE, New York, Apr. 2010, pp. 1–10.
[10] M. Nekovee, “Quantifying the availability of TV white spaces for cognitive radio operation in the UK,” in Proc. of IEEE ICC
Communications Workshops. IEEE, New York, Jun. 2009, pp. 1–5.
[11] J. van de Beek, J. Riihijärvi, A. Achtzehn, and P. Mähönen, “TV white space in Europe,” IEEE Trans. on Mobile Computing, vol. 11,
no. 2, pp. 178–188, Feb. 2012.
[12] T. Shimomura, T. Oyama, and H. Seki, “Analysis of TV white space availability in Japan,” IEICE Trans. on Communications, vol.
E97-B, no. 2, pp. 350–358, Feb. 2014.
[13] G. Naik, S. Singhal, A. Kumar, and A. Karandikar, “Quantitative assessment of TV white space in India,” in Proc. of the Twentieth
National Conference on Communications, Feb. 2014, pp. 1–6.
[14] IEEE 802.11af-2013, “Telecommunications and information exchange between systems - local and metropolitan area networks - specific
requirements - part 11: Wireless LAN medium access control (MAC) and physical layer (PHY) specifications amendment 5: TV white
spaces operation,” IEEE Standard for Information Technology, Tech. Rep., Nov. 2013.
10
[15] IEEE 802.22-2011, “Standard for cognitive wireless regional area networks (RAN) for operation in TV bands,” IEEE Standard, Tech.
Rep., Jul. 2011.
[16] IEEE 802.19.1-2014, “Telecommunications and information exchange between systems – local and metropolitan area networks – specific
requirements – part 19: TV white space coexistence methods,” IEEE Standard for Information Technology, Tech. Rep., 2014.
[17] http://www.3gpp.org/dynareport/36-series.htm.
[18] D. Gurney, G. Buchwald, L. Ecklund, S. Kuffner, and J. Grosspietsch, “Geo-location database techniques for incumbent protection in
the TV white space,” in Proc. of IEEE Symposium on Dynamic Spectrum Access Networks. IEEE, New York, Oct. 2008, pp. 1–9.
[19] S. Ghosh, G. Naik, A. Kumar, and A. Karandikar, “OpenPAWS: An open source PAWS and UHF TV white space database
implementation for india,” in Proc. of the twenty-first National Conference on Communications. NY, USA: IEEE, Feb 2015, pp.
1–6.
[20] M. Eriksson and J. van de Beek, “Is anyone out there? 5G, rural coverage and the next one billion,” IEEE Communications Technology
News, Tech. Rep., Nov. 2015.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement