The Swiss Communication Commission „ComCom“ is planning for an auction of all mobile radio frequencies in 2010/11 (including the 800 MHz band)
In this regard OFCOM held a public consultation between April and June 2009
The results of the public consultation have been published in July 2009 Reactions from Cable Network Operators: Concern is expressed that the introduction of mobile radio services in the 790-862 MHz frequency range would lead to interference in the cable networks and terminals connected to them (set-top boxes, TV sets, modems, etc.). The result would be intense interference in their service products (TV, radio, telephony and internet) for thousands of customers.
July-August 2009: OFCOM conducted measurements and theoretical studies regarding the „IMT CATV“-interference issue
On 9 th November 2009 the Federal Communication Commission „ ComCom “ instructed OFCOM to prepare the auction of mobile radio frequencies
Interference immunity level for CATV-networks and –equipment
EN 50083-8 “Cabled distribution systems for television and sound signals. Electromagnetic compatibility for networks ”
EN 50083-2 “Cable networks for television signals, sound signals and interactive services. Electromagnetic compatibility for equipment ”
CATV-networks and –equipment have a regulatory need to work up to an interference immunity level of 106 dB μ V/m (= 0.2V/m) Relevant immunity standards Background Measurements Theoretical Studies Conclusions
Radiated power of mobile handsets 1) ITU-R JTG 5-6 Doc. 5-6/88-E (Annex 14), published 3rd June 2009 user device uplink emission (e.i.r.p.) levels according to 3GPP specifications: 2 dBm EIRP = 1.6 mW 14 dBm EIRP = 25 mW 23 dBm EIRP =200 mW 24 dBm EIRP =250 mW 25 dBm EIRP =316 mW Background Measurements Theoretical Studies Conclusions The maximum radiated power of a UMTS handset is 24 dBm EIRP (as handsets have isotropic antennas the values given are in EIRP- and not in ERP) EIRP=ERP+2.15dBm ERP=EIRP–2.15dBm typical power distribution of mobile handsets data measured in an IMT rural cell in Australia 1) In this typical example 90% of the active mobile handsets have an output power of ≤ 14 dBm EIRP Accordingly 10% of the active mobile handsets have an output power of >14 dBm EIRP
Interference radius Field strength as a function of the separation distance between a LTE handset (interferer) and a CATV-receiver (victim) at a given radiation power P Tx In case a CATV-receiver complies with the immunity level of 106 d B μ V/m then the interference radius is for 90% of the active LTE handsets ≤ 4.3m If the immunity level of a device is better than 106 d B μ V/m then the interference radius is reduced accordingly Background Measurements Theoretical Studies Conclusions separation distance = interference radius victim immunity field strength P Tx interferer Remember: 24 dBm EIRP = max. radiation power of a UMTS handset 14 dBm EIRP => 90% of all active handsets radiate a power ≤ 14 dBm EIRP = regulatory interference immunity level
Measurement setups Interferer Background Measurements Theoretical Studies Conclusions Measurements have been conducted at two locations: - at a CATV-headend - in a private apartment Analogue and digital reception has been studied Separation distance between interferer and vitctim: 1.6m (to compare with studies from other organisations) UMTS signal with a bandwith of 3.84 MHz (focus on UpLink 832-862 MHz) Viewer coaxial cable analogue TV or TV set with integrated STB RF in A TV set DVB-C STB coaxial cable SCART or HDMI RF in Viewer B TV set DVB-C STB coaxial cable RF in RF out RF in coaxial cable Viewer C Cable Modem PC D coaxial cable ETHERNET CATV distribution network P Tx level increased until interference observed
Setups A & B: The immunity of the configurations under test was ≥ 10dB better than the regulatory immunity field strength limit of 106dBµV/m . The immunity could be further improved with better cables.
Setup C: When the signal was looped through the RF in /RF out connectors two devices did behave quite badly (due to built-in broadband amplifiers that pick-up interference)
Setup D: The device under test (cable modem) was a very poor one
lousy device very good device Interpretation: less than 10% of active handsets have an interference radius >1.6m
Interference cases and the results of our measurements:
insufficient shielding of the cable distribution network
Is not an issue , if the network is properly implemented
isufficient shielding of the cables between the wall socket and the receiver equipment
Critical as cable qualities differ substantially . Measured shielding effectiveness varied between 30-105dB ! => use good shielded cables only !
insufficient shielding of the receiver equipment
If a „loop-through“ of the signal via built-in broadband amplifiers was avoided: The immunity of the devices was better than the current regulatory requirement of 106dB μ V/m (exception: cable modem)
There are some CATV-devices on the market that show quite a good immunity ! => request to industry to improve the immunity of all CATV-products
If interference occurs it is mainly localised within a short distance to the victim and it is co-channel only ! Interference is thus under the control of the user of a mobile LTE handset.
TV set DVB-C STB RF in RF out RF in coaxial cable coaxial cable CATV distribution network broadband amplifier RF in RF out Warning: Daisy-chaining the signal through the RF in /RF out connectors of domestic equipment can cause a signal quality degradation up to 30dB !!
The wanted signal level in a CATV-network should be sufficient (e.g. >55 dB μ V for a 256QAM modulated OFDM-signal) to get a reasonable C/I-ratio
For use in sharing studies a realistic working level for the interfering LTE-signal is 14dBm (90 th percentile value)
Avoid daisy-chaining the signal through the RF in /RF out connectors of domestic equipment
Use only connection cables with a good shielding effectiveness
It was observed that a) interference occurs in co-channel only b) the interference radius is small (a few meters) => interference is home-made and thus under the control of the user c) the likelihood that mobile handsets radiate with high P Tx levels is relatively low what reduces the risk of interference (power distribution of LTE UE)
Criticism: In the view of OFCOM the measurements currently available from other parties are based on theoretical absolute worst case conditions (use of extreme P Tx levels for the interferer, assumption of permanent co-channel constellations and a too low wanted signal level in the cable distribution network)
Background Measurements Theoretical Studies Conclusions How likely is co-channel interference?
Probability of co-channel interference from IMT handsets (uplink) into CATV-receivers (statistical approach)
Assumptions for the theoretical modelling for Switzerland:
Active IMT handsets (UE) per km 2 and per 5 MHz: . . . . . 2.16 1) = D
Total number of CATV-receivers in SUI: . . . . . . . . . . . . . 2‘800‘000
CATV-receivers in use during busy hour: . . . . . . . . . . . . 30% (840’000)
Number of channels on the cable: . . . . . . . . . . . . . . . . . . 60
Daily usage of the CATV-receivers: .. . . . . . . . . . . . . . . . 4 h
Interference distance between UE and CATV-receiver. . <5 m = r
1) This density is conservative (i.e. is high). A large percentage of indoor mobile traffic will be operated by femto-cells in future. Femto-cells will hardly being operated in the UHF-band. Furthermore, the 3.4-GHz-band was not considered by the calculations.
Background Measurements Theoretical Studies Conclusions High cable penetration in Switzerland > 85%
Calculating the probability of co-channel interference (statistical approach)
A certain CATV-receiver is interfered , when :
this CATV-receiver is in use (active)
and an active IMT handset is located within a distance of <5 m
and the channel on the cable and the IMT channel collide
ch66 is interfered from 1 IMT channel ch67, ch68 , and ch69 are interfered from 2 IMT channels each Background Measurements Theoretical Studies Conclusions The probability that at least one IMT handset causes co-channel interference into a particular CATV-receiver is = 1.97 * 10 -5 ~ 0.002%
Number of interfered CATV-receivers (statistical approach) Distribution (CDF) of the interferences throughout Switzerland at a given point of time . Interference distance ≤ 5 m For Switzerland: With a probability of 90% less than 23 interference cases are in place at a given point of time (of a total of 2‘800‘000 devices whereas 840‘000 are in use during busy hours) Background Measurements Theoretical Studies Conclusions During main traffic hours there will be about 17 interferences on average all over Switzerland
Number of annual interferences for an individual CATV-Receiver (statistical approach) Distribution (CDF) of the number of the annually interferences per CATV-receiver and year . Interference distance ≤ 5 m Background Measurements Theoretical Studies Conclusions With a probability of nearly 90% a particular Swiss CATV-receiver will be interfered throughout one year once at most !! !
Another approach to cross-check the reliability of the statistical results („Monte Carlo“-simulation) 7 LTE cells with randomly distributed UE and CATV-Receivers
A single “shot”:
A certain CATV-Receiver is placed randomly in the centre cell
The active IMT handsets are with a given density equally positioned over all 7 cells (on the same channel)
Is there an active IMT User Equipment within a distance of <5m?
The interference from the closest UE to the chosen CATV-Receiver is calculated
Such “shots” are repeated 5 million times
Background Measurements Theoretical Studies Conclusions The supplementary „Monte Carlo“-simulation fully confirmed the results that we have previously obtained with a pure statistical investigation ! => the 90 th percentile of the mobile handset output power was identified as 13.8 dBm EIRP and the probability of co-channel interference is in the same order of magnitude ! the probability of co-channel interference is calculated