WHITE PAPER: LTE & TELEVISION SERVICES COEXISTENCE REPORT FOR BRAZIL PRODUCED BY ATDI LTD DECEMBER 2013 ON BEHALF OF GSMA

Transcription

1 ATDI Ltd The Beehive, City Place Gatwick, West Sussex RH6 0PA UK Tel: + (44) WHITE PAPER: LTE & TELEVISION SERVICES COEXISTENCE REPORT FOR BRAZIL PRODUCED BY ATDI LTD DECEMBER 2013 ON BEHALF OF GSMA 1 P a g e

2 INTRODUCTION Spectrum is limited, demand is not. The online, connected, digital world puts a strain on frequency resources that is difficult to meet. But, the notion of achieving the difficult thing was the stated motivation behind putting a man on the moon. ATDI has always shared this philosophy and believes the greatest rewards come from doing the difficult thing, in this instance making television and telephone signals coexist across one of the world s biggest countries and most prosperous economies - Brazil. OVERVIEW ATDI was asked by mobile phone and data companies in Brazil who are members of the GSM Association (GSMA) to evaluate how to use newly liberated frequencies from 703 to 803 MHz for new, long-term evolution services, known as either LTE or 4G. Specifically, GSMA members wanted to be able to demonstrate that their proposed 4G networks would coexist with current digital and analogue television signals without interference issues. The planned spectrum following refarming is shown below. Analogue and Digital TV were considered as there will be a staggered switch off and so there will still be analogue services broadcasting at the same time as new mobile services

3 BACKGROUND Measurement data exists for ISDB-T and IMT (International Mobile Telecommunications) systems and this was useful for establishing coexistence rules. However, it was also useful to determine the impact of interference which is what ATDI s study endeavoured to do. This was achieved by modelling the likely interference situation for sample areas and equating this in terms of a denial of service to a population. MODELLING APPROACH The modelling used a planning approach with planning figures based on ITU sources and data from Brazilian spectrum regulator, Anatel, to provide results that are meaningful and useful. This modelling approach has been successfully used on many similar coexistence situations in other countries to explore the issues and to inform the debate and decision process. The modelling used a number of scenarios to explore the potential interference between the broadcast and cellular services. Three study areas were used: Sao Paulo, Brasilia and Campinas. They were chosen as they are likely to be some of the most constrained areas and the potential problems should be less acute in other areas and so easier to mitigate. SCOPE AND OBJECTIVES The objective of the study was to quantify how much interference prospective LTE stations may contribute into the ISDB-T broadcasting service in terms of coverage loss for domestic receivers using a directive rooftop antenna. The impact that the digital TV services might have on the LTE base stations and user equipment (UE) was also investigated. Although the original dividend date was planned for 2016 there will now be a staggered switch off between 2015 and 2018 and so a number of analogue transmitters will continue to be operational, mainly to serve more rural population areas. Analogue TV is therefore included in this study to examine the coexistence during the transition to digital. The modelling scenarios considered are listed below LTE base station (BS) interferes with ISDB-T fixed rooftop receivers LTE user equipment (UE) interferes with ISDB-T fixed rooftop receivers ISDB-T interferes with LTE UE ISDB-T interferes with LTE BS LTE BS interfering with analogue TV LTE UE interfering with analogue TV Analogue TV interferes with LTE BS Analogue TV interferes with LTE UE - 3 -

4 METHOD The study undertook several types of modelling. The first scenario, as detailed in the previous section, used a planning tool to model the interference and the population affected. Three sample areas were considered, Sao Paulo, Brasilia and Campinas. In each area the LTE network was modelled using a representative network based on current mobile base station locations combined with typical emissive characteristics of an LTE network. For the ISDB-T network a representative network based on the existing and future stations following the implementation of the post dividend plan was used for the modelling. The remaining scenarios used the minimum coupling loss 1 approach to determine the minimum separation that would be required for each scenario so as not to suffer interference. For each scenario the analysis considered two interference mechanisms: Out of band interference - this occurs when the unwanted signals, from an adjacent channel to the receiver, are received by the receiver and interferes with the receiver preventing it from decoding the wanted signal correctly. Blocking interference - is where a strong unwanted signal prevents the receiver from detecting a wanted signal. This interference can occur even when the wanted signal is high and the effect is not as frequency selective as out of band interference. 1 ERC report

5 RESULTS The following tables give a summary of the results for each of the different scenarios. The probability of interference has been categorised into either: high, medium or low. Also shown is the probability of interference after is applied, where applicable. Mitigation options considered and those suggested for implementation are listed in the section. LTE into ISDB-T LTE BS (30m antenna) into ISDB-T (10m fixed rooftop antenna) LTE UE (1.5m) into ISDB-T (10m fixed rooftop antenna) Probability of interference Out of band emissions (modelling tool approach) Blocking (modelling tool approach) Before After Before After Medium Low Low - Low - Low - The population interfered (LTE BS into ISDB-T) is relatively low with typically less than 50,000 for Sao Paulo and less than 10,000 for Brasilia and Campinas for out of band interference. With applied those numbers could be virtually eliminated. ISDB-T into LTE ISDB-T (150m antenna) into LTE BS (30m antenna) ISDB-T (150m antenna) into LTE UE (1.5m antenna) Out of band emissions (MCL approach) Before Probability of interference After Blocking (MCL approach) before After High Low High Low Low - Low - For ISDB-T into LTE BS the separation distances required can be reduced from more than 10km to less than 600m following

6 LTE into Analogue TV LTE BS (30m antenna) into ATV (10 fixed rooftop antenna) LTE UE (1.5m antenna) into ATV (10m fixed rooftop antenna) Probability of interference Out of band emissions (MCL approach) Before After Medium Low Low - For LTE base station into analogue TV the separation distance required is typically a few kilometres. For the uplink the separation distance required is generally less than 100m. Analogue TV into LTE Probability of interference Out of band emissions (MCL approach) Before After ATV (150m antenna) into LTE BS (30m antenna) High Medium ATV (150m antenna) into LTE UE (1.5m antenna) Medium - For analogue TV into LTE base station the separation distance required is greater than 10km but reduces to below 10km with. For the LTE mobile the separation distance required is of the order of 10km

7 MITIGATION A number of techniques were considered to alleviate the potential interference issues and these are listed below. Limit emission power in top TV broadcast channels LTE filtering Broadcast filtering Use of orthogonal polarisation Domestic TV receiver filtering Improved domestic antenna Good quality TV receivers Of these options applying filters to the LTE and ISDB-T transmitter and limiting the emission power in the top TV broadcast channels are the most beneficial. Applying filters is however likely to be the most costly option but does reduce the probability of interference. The probability of interference from Analogue TV into LTE is classed as medium. Filtering on the analogue transmitters, which would reduce the out of band emissions, was not considered as the analogue transmitters are scheduled for switch off within the next few years and it is not believed to be cost effective to implement. A combination of frequency and geographic separation would be required to find an acceptable solution during the transition period from analogue to digital. The use of orthogonal polarisation was not considered viable due to the complexity and cost involved. The options for the domestic TV receiver should be used on a case by case basis where local interference issues are experienced

8 CONCLUSIONS AND RECOMMENDATIONS This report looked at the coexistence primarily between LTE and ISDB-T but also considered analogue TV effects during the transition period to a post dividend plan. There were two types of interference investigated. First, interference from adjacent channels (out of band), and second the interference through blocking. This out of band assessment concentrated on the top broadcast channels (44 to 51) and the lowest LTE uplink or downlink channel block. The blocking interference analysis considered all TV channels and not just the top channels. As a general rule the edges of the wanted service area are the most vulnerable to out of band interference and blocking is generally confined to areas in close proximity to the transmitters. For the LTE UE interfering with ISDB-T the probability of interference from blocking and out of band emissions is low. The results for the other scenarios show that out of band emissions is the dominant effect. The worst case predicted is for ISDB-T into the LTE BS where separation distances would need to be high to overcome potential blocking and out of band interference issues if no is applied. It was recommended to apply a filter to the ISDB-T transmitter to achieve, and ideally exceed the critical transmission mask. ATDI noted it would be useful to measure the resulting ISDB-T power spectra following the implementation of a filter to be able to more accurately quantify the likely improvement over the critical mask. It was also recommended to apply a filter to the LTE base station. Both of these filters combined would help to significantly lower the probability of out of band and blocking interference. There was obviously a significant cost of implementing these filters but they were projected to prove to be an effective to help coexistence. It was additionally recommended to adopt post dividend plans that avoid high power emissions in the upper TV channels (especially channels 48-51). This planning approach would help with coexistence as the potential interference problem is more acute where frequency separation to the lowest LTE band is small. The advantage of this is that it costs nothing to implement. Analogue TV will gradually be switched off and there will be instances of analogue TV transmitting in areas adjacent to LTE deployment. To avoid potential analogue TV interference caused by LTE the separation distances required are 32km (downlink) and 1km (uplink) for the co-channel case. The separation distance drops rapidly to a few kilometres (downlink) and less than 100m (uplink) for the first adjacent channel and continues to drop as frequency separation increases. The reverse case, analogue TV interfering with LTE, is likely to require higher separation distances with more than 100km required where the frequency separation is low (<10MHz). Given that the analogue transmitters will be switched off it is not considered cost effective to apply any, such as filters, to the analogue transmitters to reduce the out of band emissions. So it is recommended to have a frequency separation of at least 20MHz between the analogue TV and the LTE receiver to help reduce the separation distances required to the order of 10km. It may also be advisable to model specific cases where there are concerns about potential interference from analogue TV as the actual power and frequency separation can be taken into account and a more definitive answer provided

9 The report discusses a number of techniques and none are a solution on their own. Each technique discussed has benefits in terms of cost versus effectiveness with some considered more viable than others. Of the techniques considered filters will provide the most significant benefit but also likely to be the most expensive to universally apply. The use of orthogonal polarisation would give a benefit but is not considered as viable due to the additional cost and complexity that the mobile operators would incur. Where interference is experienced by the TV receiver there are a number of possible options such as receiver filtering, improved antenna or use of a better quality receiver. These would need to be investigated on a case by case basis to try and resolve any local interference issues but would be relatively cheap to implement. In practice a combination of the techniques was likely to be required as discussed in the report and the general conclusion was the recommendation to ensure suitable is applied so that coexistence between LTE and ISDB-T is possible. CREATION AND AUTHORSHIP The GSM Association is the professional body representing mobile phone companies, in this case those operating in Brazil. The association commissioned ATDI to create this report. It has been written by lead engineer Paul Grant who presented it at the Latin American Spectrum Conference in Brazil in the autumn of