SE24 Meeting M68
M68_19R0_SE24
Dublin, 7-8 January 2013
Date issued:
02 01 2013
Source:
Telensa Ltd, Mike Martindell
Status:
For consideration
Subject:
WI42_LTE impact on public lighting control systems
Password protected:
yes
No
x
Summary:
A report on the wide use of systems for the metering and control of public lighting operating in
the 862-870MHz SRD band. The report analyses the likely impact of LTE800 OOB emissions on
these systems.
Proposal:
For consideration, and to note the existence of this equipment and the impact of LTE
interference in the ECC/CEPT report on LTE OOB emissions and the 862-870MHz SRD band.
Background:
WI42 on LTE / SRD systems coexistence.
1
INTRODUCTION
Concerning the introduction of LTE800 adjacent to the SRD band, we would like to draw the
attention of CEPT and the SE24 committee to a substantial user of the 868 MHz SRD band of
which the committee may not be aware.
Telensa supplies public lighting control and monitoring equipment operating in this SRD band,
taking advantage of the application neutral status of the band. Telensa equipment operates using
a combination of sensitive receiver designs and Ultra Narrow Band transmissions to achieve
communication distances in cities of typically 3-4km, and further in rural areas.
There are seven million public street lights in the UK. These lights are individually controlled by
photo switches, and in many parts of the UK the local government authorities are now replacing
photo control with centralised radio control. As of December 2012, 1.4 million public street lights
have been contracted for this centralised service. Of these, Telensa has contracts for 600,000
lights, with 250,000 units already installed. Telensa expects this to rise significantly over the next
few years and that this type of equipment is likely to be installed widely across Europe.
The generic name for this type of lighting control is a Central Management System (CMS). A CMS
reports metered electricity use and faulty lights to the operator and also programs lighting to
switch off during late night to reduce energy consumption. This CMS based service is helping local
government authorities to meet EU 20-20-20 carbon reduction targets. The UK central
government provides financial support to local government to install this type of system as part of
the UK energy and carbon reduction commitments. Figure 1 shows the present coverage of
Telensa and other CMS providers in the UK. All of the public lighting control systems in use in the
UK operate in the 868 MHz SRD band.
We believe that Telensa equipment is now the largest single daily user of the SRD band in the UK.
2
COVERAGE MAP
Fig. 1: CMS Lighting Contracts with Local Government in the UK (England and Wales)
Figure 1 shows the areas where Telensa and other providers have installations or contracts to
install public lighting CMS. The grey areas are where service contracts have been bid for but the
contract winner is To Be Confirmed (TBC).
3
RADIO SYSTEM
Fig. 2: Telensa CMS system overview
This diagram shows how the Telensa system works. In a large city such as Birmingham there are
approximately 60 base stations, each has a radio connection to typically 1,000-3,000 street lights
within a 3 km radius. A Telecell is the transceiver unit that controls a street light.
Each base station has a master light sensor and the base station connects back to the central
system database using IP. The database is often connected to the local government highways
asset database and is accessed via a secure browser user interface.
4
RADIO TECHNOLOGY
The Telensa system operates with a downlink at 500mW at 869.4 – 869.65 MHz, and uplink at
20mW at 868-868.6 MHz. By use of a narrow bandwidth receiver and DSP techniques a typical
transmission distance in an urban area of 3-4 km can be achieved.
The downlink and uplink are modulated with 2FSK at 60 and 500 bits per second respectively. The
air interface uses the techniques of FEC (forward error correction), acknowledgments and retries.
Transmissions are duty cycle restricted according to the SRD band rules. LBT is not used as it
would give no advantage due to the distance between transmitter and receiver and the long
transmission frame length.
5
LTE INTERFERENCE PROFILE
The adjacent uplink band emissions of LTE800 user equipment (UE) fall within the operating band
of the Telensa system and will be seen by the receivers in the Telensa base station and Telecell as
a raised noise floor. No amount of acknowledgements, FEC or retries can overcome what is seen
by the receiver as high frequency random noise. The Telensa protocol can cope with occasional
burst interference but sustained interference will degrade reception by both the base station and
Telecell.
6
6.1
INTERFERENCE MODELLING
Interference Signal References
For our interference modelling we calculate the loss of receiver sensitivity in the Telensa units
caused by noise from a UE operating in the LTE uplink band adjacent to the SRD band.
We have used two sets of levels in our calculations and graphs.
Case 1: LTE Spurious Emissions limits
Case 1 assumes UE equipment operating to the specification limits at full power (+23dBm in
10MHz). The spurious emission limit from the ETSI TS 136 101 document at the Telensa uplink and
downlink operating frequencies is -13dBm in 1 MHz, which is -23dBm in 100kHz bandwidth (BW).
Case 2: LTE UE Measurements
Case 2 case uses the average values of the measured user equipment shown in the graph in figure
3. Figure 3 is taken from CEPT document M66_27R0_SE24 [1], which compares LTE spurious
emissions from a number of early UE devices and sources.
The figures used in the calculations for case 1 and 2 are summarised in this table.
At LTE UE transmit power of Case 1:
+23dBm / 10MHz
Spec limit
Emissions at 868.0 MHz:
Emissions at 869.4 MHz
-23
-23
Case 2:
Average values
from fig 3
-33
-46
Units
dBm in 100kHz BW
dBm in 100kHz BW
Table 1. LTE UE transmit power and spurious emission levels used in the calculations
Fig. 3: Summary of early LTE800 spurious emission level measurements in the SRD band
The measurements in figure 3 indicate that typical spurious emissions at 868MHz measure on
average at -33dBm in 100kHz. This is 10dB better than the specification limit. Note however
these measurements have been made in “ideal” laboratory set ups, where only one UE is active
and operating in a steady mode. In real life UEs will be constantly leaving and joining a base
station, resource block allocations and power control will be continually changing, and this activity
will in practice create a noisier environment than the measurements shown in Figure 3. Recent
work by APWPT [2] and TTP [3] indicate that spurious emissions do peak up to the emissions limit
line when measured over a narrow bandwidth.
6.2
Telensa base station and Telecell receiver sensitivity reduction calculations
Receiver sensitivity degradation calculations were graphed for both the base station and the
Telecell receiver using the information in table 2. The calculation steps were as follows:
1. The LTE interfering signal from the UE was scaled down from 100kHz BW figure to the
equivalent power density figure for the receiver channel noise BW.
2. The path loss from the UE to the receiver antenna was calculated for various distances at
ground level, including the polar response of the receiver antenna.
3. The path loss figure was subtracted from the calculated LTE interfering signal level to
determine the interference signal strength at the receiver
4. For the receiver, the minimum carrier to noise ratio was subtracted from the receiver
sensitivity figure to determine the threshold above which interference will affect the
receiver. For example, for the base station receiver this threshold is -143 dBm.
5. Finally the receiver threshold figure was subtracted from the interfering signal strength to
determine the loss of receiver sensitivity.
Base station
receiver
Operating Frequency
Receiver sensitivity
Minimum carrier to noise ratio
Receiver channel noise bandwidth
Height of receiver above ground
Height difference between the LTE
UE at 1m and the receiver antenna
Antenna gain
Antenna type / gain pattern
Distance between LTE interferer
and receiver
Unit
868.0
-135
8
100
9
8
Telecell (street
light
receiver)
869.4
-125
8
800
6
5
8
Co-linear
Variable
2
Monopole
Variable
dBi
MHz
dBm
dB
Hz
m
m
m
Table 2. Values used in receiver sensitivity degradation calculations
Note:
For simplicity, a free space path loss model was used for the calculations using a 20dB/decade loss
up to 300m. Over 300m a 40dB/decade path loss was applied, based on a 2-ray path loss model
for this scenario.
7
GRAPHS FROM INTERFERENCE MODELLING
These graphs show the results for modelling interference cases 1 and 2.
Fig. 4: Telensa base station performance degradation
Fig. 5: Street Light Telecell performance degradation
8
ADDITIONAL SCENARIOS
Receiver Blocking
Unlike many products in the SRD band, both ends of the link use RF bandpass filters, achieving a
receiver out of band blocking performance of typically 0dBm for the Telensa base station and
-20dBm for the Telecell. Calculations have been made for blocking performance, and we do not
expect the Telensa system to be affected by blocking either from LTE800 UE or from LTE800 base
stations.
LTE Base Station Co-location
Calculations have also been made for equipment use near a macro cell type LTE800 base station.
Provided a Telensa Telecell or base station is further than 30-50m from a macro LTE base station,
we expect few problems to occur.
9
RESULTS ANALYSIS
The system can occasionally tolerate a 10dB reduction in receiver sensitivity, although if this is a
frequent occurrence, then more conservative figures will have to be used in radio planning of
base station locations for new CMS system rollouts. The result will be a requirement for more
base stations and a higher cost to install new systems.
If interference results in a sensitivity loss of more than 10dB then there will be communication
degradation in the system and a failure of both the metering and the lighting control function of
the system.
Telensa base Station
Referring to figure 4, base station reception will be affected if a UE is closer than approximately
450m away in case 1 where one UE is transmitting at full power and spurious emissions are at the
reference LTE mask limit. A figure of 450m is of serious concern, as there will be few if any
locations for a base station in an urban area that would give protection to passing UEs, and this
figure is for just one UE. In practice there will many more UEs active within a 450m radius of one
of our base stations.
Considering case 2 where a UE is operating as the typical measured values then the situation is
improved, but still a matter of concern, with a 10dB loss of receiver sensitivity when a UE closer
than 200m to the base station.
There are a significant number of base stations already installed across the UK, some of which
may need moving, which will involve re-installation costs.
Street Light Telecell
Referring to figure 5, a street light will be affected if a UE is closer than 300m away, under the UE
maximum power conditions in case 1. Consider that in urban areas the street lights are often
immediately outside a house or apartments and much closer than 300m.
As it is impractical to relocate individual street lights, UE co-location will give significant problems
to the users of Telensa CMS, who will lose the reporting and control facilities that they expect
from the system.
In case 2 where a UE is operating at the typical measured values then the situation is significantly
improved, helped by the higher frequency of the downlink at 869.4 MHz, as the measured output
of the sample devices is typically 13dB lower than at 868 MHz. A 10dB loss of sensitivity occurs
with the UE at 20m from the street light in this case.
In summary, the Telensa radio system has been designed to high engineering standards, however
the deployment of LTE800 will have a serious affect on both existing and new installations, and
the quality of service provided.
10 CONCLUSIONS
We conclude that there is an urgent need to re-assess the spurious limits for LTE800 user
equipment, due to the emissions generated in the uplink channel. While we might take some
temporary comfort in the measurements on early LTE devices, low spurious emissions are unlikely
to be the case in the future. As has been the practice with other radio systems in the past, design
engineers will be under pressure to reduce costs and increase battery life for both LTE800 chipset
designs and products. Filtering will be minimised to save cost and space, and PAs will be operated
at the edge of linearity to increase battery life, with subsequent increases in spurious emissions
up to the limit line.
In the current LTE specifications, the spectral mask treats the SRD band the same as an adjacent
LTE channel, but the SRD band must be considered as a separate radio band. The ITU radio
regulations for adjacent band interference must be considered. The level of spurious LTE800
emissions in the SRD band should be reduced to a level consistent with the adjacent band
spurious / OOB limits applied to other communication systems using the UHF band.
11 APPENDIX: OTHER CMS SYSTEMS
There are four other CMS suppliers that use the same SRD band, but based on a shorter range
mesh network topology. The technical characteristics of these systems are not known. We cannot
comment on the performance of these systems in the presence of LTE800 signals, except to say
that as with other devices in the SRD band, performance problems are likely under conditions
similar to those described in this paper.
12 REFERENCES
[1] CEPT M66_27R0_SE24_att_Level comparison2, 17/10/12
[2] CEPT SE24 M65_29R0 APWPT, A study of LTE interference potential with regard to PMSE
operation, 18/06/12
[3] UK ICTKTN / Cambridge Wireless Radio Technology SIG paper, The Elephant in the Radio, TTP,
08/12/12