With so many different possibilities and options available to the operators for deployments it is necessary that the planning is done in such a way as to keep the interference to minimum and throughput to maximum. Let’s look at two simple deployment situations. For simplicity let’s assume that the operator has won 20MHz of contiguous spectrum. We have 5 UE’s. UE#1, UE2 and UE#5 are basic Rel-8 LTE UE’s and UE#3 and UE#4 are LTE-A UE’s with support for Carrier Aggregation (CA) and enhanced Inter-Cell Interference Coordination (eICIC) Operator 1 decides to split the 20MHz spectrum into 15MHz for macrocell and 5MHz for small cell. In this case: UE#1 and UE#2 get 5MHz of spectrum UE#3 gets 15MHz + 5MHz using CA. Total 20MHz UE#4 and UE#5 get 15MHz of spectrum. The advantage of this scenario is a simple Hierarchical Cell Structure (HCS) deployment. There is no need to worry about the interference. Operator 2 decides to deploy macrocell and small cells in Heterogeneous Network (HetNet) deployment (co-channel). They have used the complete 20MHz channel for both the Macrocells and Small cells. Due to the interference experienced by the UE’s on small cells, the Macrocell is using ABS (Almost Blank Subframes). The advantage of using ABS is that it can be dynamically configured. Let’s consider a peak and non-peak scenario from macrocell point of view. In the peak scenario macrocell is having 10% ABS and in non-peak it is having 50% ABS. Let’s look at the user impact Non-Peak scenario (50% ABS) UE#1 – 20MHz (Since it’s in the center of the cell, we will assume no interference from Macrocell) UE#2 and UE#3 – 20MHz (High interference in 50% of the time) UE#4 and UE#5 – 20MHz (ABS for 50%) In this situation, the throughput for UE#4 and UE#5 would be half or equivalent of 10MHz scenario. The throughput for UE#2 and UE#3 would equivalent to somewhere in between 10MHz and 20MHz scenario. Peak scenario (10% ABS) UE#1 – 20MHz (Since it’s in the center of the cell, we will assume no interference from Macrocell) UE#2 and UE#3 – 20MHz (High interference in 90% of the time) UE#4 and UE#5 – 20MHz (ABS for 10%) In this situation, the throughput for UE#4 and UE#5 would be or equivalent of 18MHz scenario. The throughput for UE#2 and UE#3 would equivalent to somewhere in between 2MHz and 20MHz scenario. Even though it may be advantageous to have the deployment as proposed by Operator 2, interference may cause more problems in the long term. I am interested in hearing what option do you think will work best and why. ... View more

It’s happening again! The excitement, business cases, discussion on how the technology has matured, lessons learnt from previous such rollouts, etc. Believe it or not, it’s happening all over again. LTE Broadcast TV (a.k.a. eMBMS) is coming to an operator near you, soon. Back in 2006, when Release-6 of UMTS was released, MBMS (without the leading ‘e’) was being hailed as a great technology that would solve many of the ills that had been plaguing the Mobile TV rollout. For example, the biggest issue was additional spectrum that was required with any of the other Mobile TV Broadcast technology, was not a problem for MBMS. In case of MBMS (Multimedia Broadcast Multicast Service), the spectrum of the UMTS channel (fixed 5MHz) could be dynamically partitioned to serve the regular Voice(CS) + Data(PS) traffic and the broadcast data. None of the other competing broadcast standards then like DVB-H, T-DMB, ISDB-T, CMMB and MediaFLO could offer such an advantage. Another big advantage with having 3GPP cellular broadcast standard (MBMS) in comparison to the competing technologies was that no additional hardware/chipset was required and there was no necessity for additional authentication and security mechanisms. Even after many such advantages, MBMS never got off the ground. The simplest of explanations revolved around the limitation that UMTS channels bandwidth is fixed to 5MHz, which means only limited number of channels could be supported for Mobile TV transmission. Another reason was that the operators tried to do too much too soon and as a result their business case fell flat. This was a result of using Multicast to sell subscription services to the users who had very little or no experience of watching TV/Video. Let’s look at the broadcast and multicast concept in detail. Unicast, Broadcast and Multicast In case of ‘Unicast’, the radio access network (RAN) has to setup a dedicated bearer with the cellular device and then transmit the broadcast video. This would defeat the purpose of broadcast as a dedicated bearer is set up with the device and the device is effectively using the data. This is not a preferred approach and used in extreme cases for the sake of continuity. If only a few users in the cell are watching the mobile TV then there could be a saving of bandwidth by letting each of these users have a unicast connection rather than sending all information using the broadcast. Unicast mode is also known as ‘one-to-one’ or ‘point-to-point’ (ptp) transmission. Normal video streaming (using Youtube, Netflix, etc.) is always using the Unicast mode. In case of ‘Broadcast’ mode, the transmitted information is available for every device to be able to view. Broadcast mode is also known as ‘one-to-many’ and ‘point-to-multipoint’ (ptm) transmission. ‘Multicast’ mode is a special case of Broadcast mode where the information may be available for all users but could only be decoded / deciphered by a device that belongs to the multicast group. To belong to this group, the user would have to subscribe to the service beforehand by calling the operator or using some online website, etc. While in case of 3G MBMS, all the three modes were supported, in case of LTE eMBMS (‘e’ stands for evolved), Multicast mode is not supported. To highlight the similarity with 3G MBMS, the abbreviation was not changed to eMBS. High profile Mobile TV launches in the past Over the last few years, many big players have tried their hands on Mobile TV. Here is a summary of a few of them: MediaFLO : A very ambitious and bold Mobile TV attempt was made by Qualcomm when it launched its services back in June 2009. Initially it was sold by AT&T and Verizon but the users had to pay $15 for subscription per month. This pricing was reduced and there were also other discounts available for users to sign up to the service. Qualcomm also sold a standalone device with subscription and tried to partner for in-car entertainment systems. The main reason for failure was high subscription prices for limited content and lack of smartphone models supporting MediaFLO. We have to remember that this required additional spectrum and hardware (chipset) which meant additional subscription charges. This service was eventually shut down in early 2011. NOTTV : Japan has always been a trendsetter and a leader in technology. No discussion on Mobile TV could be complete without mentioning Japan or their leading operator NTT Docomo. Back in April last year, they announced that they have 680K subscribers to their NOTTV Mobile TV service after a year of launch (though they were expecting atleast 1 million). Each subscriber pays 420JPY (roughly $4/£2.5/€3) per month. One of the ways NOTTV was made appealing to the end subscibers was by providing original content that was only available here and was also archived so playback was possible too. Subscribers can also provide live feedback or answers to what was being shown thereby increasing participation and value over the traditional television. China Mobile TV Service : China Mobile is another operator with clout and loads of subscribers. It has been pushing the Chinese mobile TV standard (CMMB – China multimedia mobile broadcasting), not only in China but in other parts of the world as well. Again, this requires an additional hardware and spectrum for the receivers to be able to receive the content. A report back from 2010 suggested that the number of users of this service were much less than expected and only a few of them were actually paying subscribers. China Mobile Hong Kong launched mobile TV services based on CMMB in Dec. 2011. CMMB based mobile TV is also being launched in Philippines this year . Many other operators and other television & media companies have launched mobile TV services based on the streaming (unicast) model discussed above. While this may work in the short term, in the long term this is going to congest the mobile networks thereby impacting the traditional voice and data services. An easy option available with the operators is reduce the priority of the mobile TV data but this would mean the quality of experience (QoE) of the mobile TV subscribers would suffer and they may desert the services. ‘eMBMS’ as the saviour Back in March last year , a top Verizon executive confirmed that they will be launching Mobile TV based on LTE broadcast technology, eMBMS, sometime in 2014. In June last year, Verizon is reported to have agreed a multiyear $1 billion deal with NFL for the rights to broadcast the games on smartphones. The deal though is only for the smartphones, not for the tablets. My guess is that it’s for any device that has a SIM card in it. eMBMS would make sense for broadcasting content such as live games to a wide audience without overloading the network. AT&T doesn’t want to be left behind and its building its own eMBMS network on the old MediaFLO spectrum it bought off Qualcomm. In fact, if it reserves an entire 5MHz spectrum available nationally for eMBMS, it can use the alternative eMBMS configuration of 7.5KHz channels (rather than the regular 15Khz channels) which could result in more channels being available and also better performance. Finally, the Australian operator Telstra recently conducted LTE-Broadcast (eMBMS) trials over its commercial 4G network, broadcasting several sport events and even a file download to several mobile devices over the same wireless transmission. Qualcomm and Ericsson, who partnered Telstra in these trials, believe that they have found the right model to make broadcasting work. Do users want Mobile TV The short answer is, of course they do. I remember being told many years back about this survey where the users were asked if they would want TV on their mobile and if they would prefer to pay for that. The answer was a resounding yes. The only problem with that survey was that nobody asked the respondents what they understood by Mobile TV and how much would they prefer to pay. Over the last many years I remember asking people I meet in various works of life the same questions. The most common answers I get are; Mobile TV is like Youtube or iPlayer and the maximum about anyone would prefer to pay is £2($3). I am sure this is not what the operators expect. In fact in this day and age where the Freemium model is being used for Apps and services, are the users not going to expect the same from any Mobile TV offering. Maybe some users wouldn’t mind paying extra in a bundle offering. The above picture from the Adobe’s digital Index team highlights the important point that users still prefer watching video on tablets, rather than the small smartphone screens. This picture above from Business Insider article early last year highlights the difference in viewing habits with smartphone and other kind of devices. Frankly, I am surprised by the number of users on the smartphone watching video longer than 10 minutes. Another piece of statistics from an eMarketer article, also from early last year, shows that the top three kinds of content for both smartphones and tablet users were movies, user-generated content (such as YouTube videos) and TV shows. But the difference lies in emphasis: Tablet viewers were much more likely than mobile phone viewers to prefer feature-length movies and TV shows. Mobile phone viewers were more likely to watch user-generated content. It is important to highlight that the span of attention and the patience required watching lengthy content on smartphone is a tricky job. Mobile TV is exactly what smartphone users don’t want. There’s still hope for eMBMS and Mobile TV I have tried my best to reason why Mobile TV on smartphone may be difficult to succeed. Tablets are becoming increasingly the main means of watching lengthy videos but most of them are Wi-Fi only. Two simple ways in which Mobile TV uptake may get a boost would be to have unique content, tailored for smaller screens and to have similar content being broadcasted on other connected devices like tablets, regardless of whether they are Wi-Fi only or support cellular access. Without allowing these alternative devices to receive Mobile TV, eMBMS may suffer the same fate as those of MBMS and MediaFLO. About the Author Read other blog posts by Zahid Ghadialy here Follow Zahid on Twitter @zahidtg ... View more

Download the full slidecast below. For all other blog posts and content by the author, Zahid Ghadialy, visit his community profile. ... View more

A few questions are being asked recently regarding the impact of M2M on Small Cells. With a prediction of up to 50 billion devices by 2020-25, this is a very valid question. While M2M applies to a broad range of devices, there are still people in the industry who tend to think of M2M devices as small, cheap devices, generally with little or no mobility, responsible for small amounts of data transfer over long durations. While it is true that a multitude of M2M devices may fall in this category, it is not necessarily applicable for all M2M devices. For instance, the car industry is working very tough on the design and development of various ‘connected car’ initiatives. These connected cars will require high speed data transfer (e.g. SatNav maps, traffic related information, video streaming, etc.) over a fast varying channel due to very high mobility. To facilitate and simplify this analysis an assumption of little or no mobility will be made where device ‘may’ or ‘may not’ require reasonable amount of data at regular time intervals. This assumption would also be more applicable to scenarios with possible small cells around. Such devices could be motion sensor based devices like security cameras, connected meters, any other sensors that monitor real time information, etc. We can discount M2M devices like those in connected cars from our analysis because they would most likely rely on macro cellular coverage. Finally, we will consider only ‘open access’ small cells rather than ‘closed access’ (a.k.a. CSG or ‘Closed Subscriber Group’) small cells. There are many factors that need to be considered in the design of an M2M device such as cost, form factor, power consumption, security, etc. To keep the cost down while maintaining a small and simple form factor, it may be sensible to stick with one access technology rather than more. Power consumption is very low in certain type of technologies like Bluetooth; it is not always practical for large number of M2M devices. Each of these M2M devices may need a Bluetooth access point which may not always be feasible. On the personal front, many eHealth M2M devices use Bluetooth as access technology where the user provides some sort of input and the M2M device can connect to the Bluetooth on the phone. Generally, most indoor low mobility M2M devices would rather have wireless coverage which is WiFi or cellular. A point generally made in favor of WiFi is that it’s cheap but as most people would already know, ‘there is no such thing as a free lunch’ and there are other issues that need to be considered along with this. First and foremost being that WiFi coverage area is limited per access point which may be true in case of cellular small cells also. However the M2M device can always fall back on the macro cell usage in case if the small cell becomes unavailable for whatever reason. WiFi makes use of unlicensed spectrum which is prone to interference and jamming issues. Another issue that needs consideration is if the WiFi channel is security protected or not. If unsecured, the data sent by the M2M devices may be visible for others to view unless the M2M devices encrypt it thereby increasing complexity and maybe cost. Security of the devices may also be compromised if some kind of vulnerability is detected after the devices are in the field when using WiFi. With cellular, these issues are much reduced as the SIM provides the additional layer of security against the potential hackers and also the user data is not visible for anyone interested in eavesdropping or sniffing. In case of security protected WiFi devices each of the M2M devices would need to possess the appropriate security credentials. If the WiFi SSID changes or password is changed then each of these devices would have to somehow update their credentials. With cellular, the device relies on SIM for authentication and security and thereby benefits both the network as well as the device as they both know that the other party is a trusted one. With WiFi out of consideration, an obvious question would be how small cells could do the job better than the macro cell? The ground reality is that the macro cells are quite heavily in use most of the day. There is no longer a ‘peak period’ but the use is generally distributed evenly during an entire day. The standardization bodies along with the operators are working on various ways to offload the traffic from the macro cells to some other form of access networks to make sure there is cell capacity available for users who cannot be or would rather not offload. Going back to the predictions of 50 Billion devices, there is a limit on how many active users the network can simultaneously allow on a cell. The restriction can arise because of the ‘air interface’ bottlenecks or even the ‘core network’ overload. The networks are also wary of ‘Signalling Tsunamis’ which are very much possible with a multitude of M2M devices. Some of the new features, especially in LTE, deal with access barring of M2M devices to avoid overload. Small cells with a restricted coverage area may be less prone to the ‘overload’ situations in the access or/and the core network. There are also features intended for small cells that will allow the user plane data to be offloaded while the signaling data, responsible for security, is still sent through the normal route. The above mentioned and some additional features are still under development and standardization process by the 3GPP which will work to the benefit of Small Cells thereby making them the best solution for certain types of M2M devices that we have considered in this analysis. ... View more