Updates from the 3GPP RAN 5G Workshop - Part 2

I have finally got round to having a look at some more presentations on 5G from the recently concluded 3GPP RAN 5G Workshop. Part 1 of the series is here.

Panasonic introduced this concept of Sub-RAT's and Cradle-RAT's. I think it should be obvious from the picture above what they mean but you can refer to their presentation here for more details.

Ericsson has provided a very detailed presentation (but I assume a lot of slides are backup slides, only for reference). They have introduced what they call as "NX" (No compatibility constraints). This is in line to what other vendors have referred to as well that above 6GHz, for efficiency, new frame structures and waveforms would serve best. Their slides are here.



Nokia's proposal is that in the phase 1 of 5G, the 5G Access point (or 5G NodeB) would connect to the 4G Evolved Packet Core (EPC). In phase 2, both the LTE and the 5G (e)NodeB's would connect to the 5G core. Their presentation is available here.

Before we move on to the next one, I should mention that I am aware of some research that is underway, mostly by universities where they are exploring an architecture without a centralised core. The core network functionality would be distributed and some of the important data would be cached on the edge. There will be challenges to solve regarding handovers and roaming; also privacy and security issues in the latter case.

I quite like the presentation by GM research about 5G in connected cars. They make a very valid point that "Smartphones and Vehicles are similar but not the same. The presentation is embedded below.

Qualcomm presented a very technical presentation as always, highlighting that they are thinking about various future scenarios. The picture above, about phasing is in a way similar to the Ericsson picture. It also highlights what we saw in part 1, that mmW will arrive after WRC-19, in R16. Full presentation here.

The final presentation we are looking is by Mitsubishi. Their focus is on Massive MIMO which may become a necessity at higher frequencies. As the frequency goes higher, the coverage goes down. To increase the coverage area, beamforming can be used. The more the antennas, the more focused the beam could be. They have also proposed the use of SC-FDMA in DL. Their presentation is here and also embedded below.

4G Mobile Network & Applications

4G Mobile Network & Applications

View more presentations from Kuncoro Wastuwibowo.

OFDMA Femtocells: A Roadmap on Interference Avoidance

Earlier, I have blogged about LTE femtocells being starting point of LTE and how LTE can be better technology than HSPA. In this months IEEE Communications magazine, there is a series of articles on Femtocells. I will try and cover some of these (unless I wander off in some other direction). The first one is titled 'OFDMA Femtocells: A Roadmap on Interference Avoidance'. At the end of this post, I have provided links to the research and the actual paper (in a legal way ;) so if you are not interested in the post and want to directly jump on the actual paper see the end of this post.

There are all kinds of statistics about the number of Femtocells worldwide. There could be upto 70million by 2012. If this happens the big problem would be the interference between Macro and Femtocells and also between Femtos. OFDMA (used in LTE and WiMAX both) Femtocells can handle the interference better than CDMA (UMTS and CDMA2000) Femtocells due to its Intracell interference avoiding properties and robustness to multipath.

So what are the main problems that the operators will face when deploying femtocells? Lets look at some of them:

• Access method: Three different approaches exist namely, Open access, Closed access and Hybrid access which is a mix of both of them. The first two approach has some problems and I have suggested a solution before ;) but the best solution may be to go for Hybrid approach where limited connectivity is available to non-subscribers of the femto.
• Time Synchronisation is another important aspect of OFDMA Femtos. To minimise multi-access interference and for successful handovers, synchronisation between all the Femtos and between Femto and Macro is a must. This should be acheived without any complicated hardware so as to keep the cost down.
• Physical Cell Idendities (PCI) could be a problem because of limited numbers
• Neighbouring cell list, which is restricted to 32 in LTE, could be a problem if too many Femtos are around
• Handovers could also be a problem if the UE keeps jumping between Femtos and macro. One solution could be the use of HCS.

Interference analysis will definitelty play an important part in the rollouts. If not properly managed, could result in dead zones within Macro. Power control Algorithms and Radio Resource Management strategy will help but effective Spectrum allocation technique is needed as well. The diagram above shows different approaches for subchannel allocation in OFDMA femtocells.


The Femtocells would need to be self-configurable and self-optimising. I tried to explain the SON concept earlier which is similar. Self-configuration comes into picture when the Femto is switched on. Once the parameters are adjusted then Self-Optimisation tries to optimise these defaults into something better and more suited to the current environment. Sensing of the environment plays an important part in this. The diagram above shows different approaches being used by different Femtocells. The cheapest approach would ofcourse be the measurement report approach where the phone is made to report the environment. The only problem being that whichever phone was used (automatically selected) will have considerable amount of its battery power used up :)

The team behind this IEEE paper has been doing some excellent research work in the field of femtocells.

There is a book that is under publication and will be available early next year. At the same time if it interests you, you can look at some of their publications including the IEEE one that has been quoted here. Here are all the necessary links:

• The Centre for Wireless Network Design
• Professor Jie Zhang homepage
• David López-Pérez, Research Fellow, homepage
• Femtocells World Summit presentation on Self-organisation of the Subchannel Assignment in OFDMA Femtocells
• OFDMA Femtocells: A Roadmap on Interference Avoidance - IEEE Communications magazine, Sep. 2009
• ICC 2009: Femtocells panel presentation slides

Hope someone finds all this info useful :)

Inter symbol and inter carrier interference (ISI and ICI) in OFDMA

Radio channel are random, fast changing and error prone. In a wireless system the variation/fluctuation in the received signal is called fading. The goal of the wireless system design is to overcome different types fading and provide reliable and efficient transmission. Generally there are two types of fading.

• Large scale fading: It is the fluctuation in the average signal strength over a large distance and is caused by terrestrial change. This occurs when a mobile travel from a lake to mountainous are to a lake area or from an open area to a tall buildings area. Large scale fading can be mitigated by controlling the transmit power.
• Small scale fading: Occurs as a result of the fluctuations in the received signal strength over a small distance and is caused by multipath and Doppler's shift. Doppler shift refers to the change on frequency of the signal because of relative motion between the transmitter and the receiver.

The figure here shows the multipath propagation for a signal. Signal goes from transmitter to the receiver through multipath that have different lengths i.e. path 1, path2 and path 3. The signal from different path arrives at the receiver at different times although it’s originated from the same source. The received symbol as shown below is longer than the duration of the original symbol.

Delay spread can cause adjacent symbols to interfere at the receiver. As a result of the multipath the delayed version of the first symbol shifts into the next symbol time and thus causes overlap between he symbols. In OFDMA this is taken care of where more time is give for each symbol to be received at the receiver by inserting a guard time.

The Doppler shift introduces another type of interference in OFDMA i.e. inter carrier interference (ICI). OFDMA divided the spectrum into narrowband subcarriers and they are tightly spaced simply because they are orthogonal. One of the requirements for orthogonality is to maintain the subcarrier spacing exactly the reciprocal of the symbol period. The figure below shows the frequency shifts thus changing the subcarrier spacing which results in the loss of orthogonality. This loss of orthogonality creates interference among the signals which is called as ICI. Since the subcarriers in OFDMA are usually very narrow hence the OFDMA system becomes very sensitive to ICI. ICI destroys the orthogonality of the OFDMA system which is overcome by the use of cyclic prefix mechanism.

Under this mechanism OFDM symbols are extended into periodical symbols i.e. redundant information is sent out to ensure that analysis can be conducted on the undistorted information and is called as cyclic extension.

It can be implemented by copying the portion of the original symbol from the end and attaching it to the front or copying it from the front and attaching it to the end. Since OFDMA has already assigned the guard time to defeat ISI, cyclic extension can be put into the guard time interval. This is called cyclic prefix. With cyclic prefix used the delayed version of the previous symbols cannot shift into the useful time of the current symbol so ISI is eliminated as well. Also the cyclic prefix provides redundant information and allows spectral analysis in the receiver to maintain the orthogonality of the subcarriers. Thus the cyclic prefix can be used to deal with both ISI and ICI.

The above concepts can be summarized in teh form of the picture below

OFDM and SC-FDMA

OFDM has been around since the mid 1960s and is now used in a number of non-cellular wireless systems such as Digital Video Broadcast (DVB), Digital Audio Broadcast (DAB), Asymmetric Digital Subscriber Line (ADSL) and some of the 802.11 family of Wi-Fi standards. OFDM’s adoption into mobile wireless has been delayed for two main reasons. The first is the sheer processing power which is required to perform the necessary FFT operations. However, the continuing advance of signal processing technology means that this is no longer a reason to avoid OFDM, and it now forms the basis of the LTE downlink. The other reason OFDM has been avoided in mobile systems is the very high peak to average ratio (PAR) signals it creates due to the parallel transmission of many hundreds of closely-spaced subcarriers. For mobile devices this high PAR is problematic for both power amplifier design and battery consumption, and it is this concern which led 3GPP to develop the new SC-FDMA transmission scheme.

The LTE downlink transmission scheme is based on OFDM. OFDM is an attractive downlink transmission scheme for several reasons. Due to the relatively long OFDM symbol time in combination with a cyclic prefix, OFDM provides a high degree of robustness against channel frequency selectivity. Although signal corruption due to a frequency-selective channel can, in principle, be handled by equalization at the receiver side, the complexity of the equalization starts to become unattractively high for implementation in a mobile terminal at bandwidths above 5 MHz. Therefore, OFDM with its inherent robustness to frequency-selective fading is attractive for the downlink, especially when combined with spatial multiplexing.

Additional benefits with OFDM include:
• OFDM provides access to the frequency domain, thereby enabling an additional degree of freedom to the channel-dependent scheduler compared to HSPA.
• Flexible bandwidth allocations are easily supported by OFDM, at least from a baseband perspective, by varying the number of OFDM subcarriers used for transmission. Note, however, that support of multiple spectrum allocations also require flexible RF filtering, an operation to which the exact transmission scheme is irrelevant. Nevertheless, maintaining the same baseband-processing structure, regardless of the bandwidth, eases the terminal implementation.
• Broadcast/multicast transmission, where the same information is transmitted from multiple base stations, is straightforward with OFDM.

For the LTE uplink, single-carrier transmission based on DFT-spread OFDM (DFTS-OFDM) is used. The use of single-carrier modulation in the uplink is motivated by the lower peak-to-average ratio of the transmitted signal compared to multi-carrier transmission such as OFDM. The smaller the peak-to-average ratio of the transmitted signal, the higher the average transmission power can be for a given power amplifier. Single-carrier transmission therefore allows for more efficient usage of the power amplifier, which translates into an increased coverage. This is especially important for the power-limited terminal. At the same time, the equalization required to handle corruption of the single-carrier signal due to frequency-selective fading is less of an issue in the uplink due to fewer restrictions in signal-processing resources at the base station compared to the mobile terminal.

In contrast to the non-orthogonal WCDMA/HSPA uplink, which also is based on single-carrier transmission, the uplink in LTE is based on orthogonal separation of users in time and frequency. Orthogonal user separation is in many cases beneficial as it avoids intra-cell interference. However allocating a very large instantaneous bandwidth resource to a single user is not an efficient strategy in situations where the data rate mainly is limited by the transmission power rather than the bandwidth. In such situations, a terminal is typically allocated only a part of the total transmission bandwidth and other terminals can transmit in parallel on the remaining part of the spectrum. Thus, as the LTE uplink contains a frequency-domain multiple-access component, the LTE uplink transmission scheme is sometimes also referred to as Single-Carrier FDMA (SC-FDMA).

Via: 'Agilent Whitepaper' and '3G evolution'.

MIMO schemes in LTE

SU-MIMO (Single User MIMO)

•This is an example of downlink 2x2 single user MIMO with precoding.

•Two data streams are mixed (precoded) to best match the channel conditions.

•The receiver reconstructs the original streams resulting in increased single-user data rates and corresponding increase in cell capacity.

•2x2 SU-MIMO is mandatory for the downlink and optional for the uplink

MU-MIMO (Multiuser MIMO)

•Example of uplink 2x2 MU-MIMO.

•In multiple user MIMO the data streams come from different UE.

•There is no possibility to do precoding since the UE are not connected but the wider TX antenna spacing gives better de-correlation in the channel.

•Cell capacity increases but not the single user data rate.

•The key advantage of MU-MIMO over SU-MIMO is that the cell capacity increase can be had without the increased cost and battery drain of two UE transmitters.

•MU-MIMO is more complicated to schedule than SU-MIMO

Everything you want to know on Single Carrier FDMA

While working on our LTE training, I came across this very interesting website that contains probably everything you want to know on SC-FDMA. Bookmark it if this is an area of interest.

Single Carrier FDMA Discussion Forum

SDWN: Beyond Femtocells

Missed this one earlier from Fierce Broadband Wireless:

Femtocells have little ability to become self-organized or perform network management functions. WiMAX and LTE, on the other hand, are based on highly adaptive OFDMA air interfaces and IP communications that enable architecting of self-configured and distributed networks. They become part of the broader unified communications industry movement toward smart, distributed networking that is augmented by distributed storage and application servers. The largest opportunities for WiMAX are Smart Distributed WBB Networks, SDWN, and Purpose Use of Multiple Spectrum bands (PUMS).

The SDWN wireless interface layer is based on scalable OFDMA, adaptive modulation and power control methods that enable the network to adapt to a variety of channel bandwidths, range, multi-path environments, and attenuating signal conditions and usage.

The same factors that are now driving intelligent wired networks are amplified by the nature of wireless: limited available spectrum. MIMO- beam forming, smart antenna and smart power regulation built into remote and mobile stations can achieve very high localized performance while working as an adaptive layer within the managed network.

The term ‘WiMAXmesh' is used by some companies to describe SDWN functionality for multi-hop relay network capabilities now being implemented into WiMAX integrated circuits and network devices. Currently, WiMAXmesh can include on-board or local memory and storage, support for multiple external network interfaces and optimized network routing capabilities.

SDWN is highly motivated by the need to deliver cost effective networks able to respond to high bandwidth demands from both enterprise and consumer markets. The need to be integrated as part of enterprise level networks compels development of self-configuring, self-organizing WBB networks. The growing demand for personal broadband, including social networking and video media, will propel the pace of SDWN developments.

We believe that SDWN will develop over the next 10 years to become two to four times larger than the conventional centralized wireless broadband network market.

Key stake holders in SDWN development include most major companies involved in WiMAX and 3G-LTE including:

• Alcatel-Lucent: A leading developer of co-MIMO, MU-MIMO technologies
• Alvarion: An early implementer of distributed network capabilities
• Cisco: We think that Cisco intends to become the leader in SDWN for both WiMAX and LTE. Products may not appear for up to two years.
• Intel: Offers distributed processor architectures and enabling IP.
• Nortel: An early leader in MIMO-OFDM and is continuing development toward SDWN
• Motorola: The company is not as visible but has developed corresponding IP
• Ericsson: Has recently entered submittals to 802.16 that correspond to work in LTE.
• Huawei: A rising developer in SDWN technologies.
• Numerous efforts are underway among WiMAX, LTE and multi-mode chip and smaller equipment suppliers. picoChip and DesignArt are noticeable examples.

Keywords:

• WBB - Wireless Broadband
• WBBN - WBB Networks
• SDWN - Smart Distributed WBB Networks
• PUMS - Purpose Use of Multiple Spectrum bands

On OFDMA ...

Last year I wrote a blog on the difference between OFDM and OFDMA. Lots of people found it useful.

With OFDM/OFDMA being discussed everywhere, I thought of linking an article which would give everyone quick info on these technologies. This article from Network Systems Designline titled "Understanding OFDMA, the interface for 4G wireless" is split into two parts:

1st is Introduction to OFDM

2nd is on OFDMA

Let me know if you found these of use.

OFDM and OFDMA: The Difference

I was curious as to why IEEE 802.16d (fixed service) uses Orthogonal Frequency Division Multiplexing (OFDM). IEEE 802.16e (mobile) uses Orthogonal Frequency Division Multiple Access (OFDMA). So, what’s the difference between the two, and why is there a difference?

Lets first look at FDM:

In FDM system, signals from multiple transmitters are transmitted simultaneously (at the same time slot) over multiple frequencies. Each frequency range (sub-carrier) is modulated separately by different data stream and a spacing (guard band) is placed between sub-carriers to avoid signal overlap.

OFDM is sometimes referred to as discrete multi-tone modulation because, instead of a single carrier being modulated, a large number of evenly spaced subcarriers are modulated using some m-ary of QAM. This is a spread-spectrum technique that increases the efficiency of data communications by increasing data throughput because there are more carriers to modulate. In addition, problems with multi-path signal cancellation and spectral interference are greatly reduced by selectively modulating the “clear” carriers or ignoring carriers with high bit-rate errors.

Like FDM, OFDM also uses multiple sub-carriers but the sub-carriers are closely spaced to each other without causing interference, removing guard bands between adjacent sub-carriers. This is possible because the frequencies (sub-carriers) are orthogonal, meaning the peak of one sub-carrier coincides with the null of an adjacent sub-carrier.

In an OFDM system, a very high rate data stream is divided into multiple parallel low rate data streams. Each smaller data stream is then mapped to individual data sub-carrier and modulated using some sorts of PSK (Phase Shift Keying) or QAM (Quadrature Amplitude Modulation). i.e. BPSK, QPSK, 16-QAM, 64-QAM.

OFDM needs less bandwidth than FDM to carry the same amount of information which translates to higher spectral efficiency. Besides a high spectral efficiency, an OFDM system such as WiMAX is more resilient in NLOS environment. It can efficiently overcome interference and frequency-selective fading caused by multipath because equalizing is done on a subset of sub-carriers instead of a single broader carrier. The effect of ISI (Inter Symbol Interference) is suppressed by virtue of a longer symbol period of the parallel OFDM sub-carriers than a single carrier system and the use of a cyclic prefix (CP).

The OFDM spread-spectrum scheme is used for many broadly used applications, including digital TV broadcasting in Australia, Japan and Europe; digital audio broadcasting in Europe; Asynchronous Digital Subscriber Line (ADSL) modems and wireless networking worldwide (IEEE 802.11a/g).

Like OFDM, OFDMA employs multiple closely spaced sub-carriers, but the sub-carriers are divided into groups of sub-carriers. Each group is named a sub-channel. The sub-carriers that form a sub-channel need not be adjacent. In the downlink, a sub-channel may be intended for different receivers. In the uplink, a transmitter may be assigned one or more sub-channels.

Subchannelization defines sub-channels that can be allocated to subscriber stations (SSs) depending on their channel conditions and data requirements. Using subchannelization, within the same time slot a Mobile WiMAX Base Station (BS) can allocate more transmit power to user devices (SSs) with lower SNR (Signal-to-Noise Ratio), and less power to user devices with higher SNR. Subchannelization also enables the BS to allocate higher power to sub-channels assigned to indoor SSs resulting in better in-building coverage.

Subchannelization in the uplink can save a user device transmit power because it can concentrate power only on certain sub-channel(s) allocated to it. This power-saving feature is particularly useful for battery-powered user devices, the likely case in Mobile WiMAX.

The WiMAX forum established that, initially, OFDM-256 will be used for fixed-service 802.16d (2004). It is referred to as the OFDM 256 FFT Mode, which means there are 256 subcarriers available for use in a single channel. Multiple access on one channel is accomplished using TDMA. Alternatively, FDMA may be used.

On the other hand, OFDMA 128/512/1024/2048 FFT Modes have been proposed for IEEE 802.16e (mobile service). OFDMA 1024 FFT matches that of Korea’s WiBRO. OFDM 256 also is supported for compatibility with IEEE 802.16d (fixed, 2004).

Qualcomm, OFDM and 4G (17/05/07)

Qualcomm is the pioneer of next generation wireless technologies. To stregthen their position further, they have also bought over some smaller companies to give them access to all their IPR, etc. Yesterday i read an interview of Bill Davidson, senior vice president of investor relations and international marketing at Qualcomm and IDG news service. Here are some interesting points:

IDGNS: Is OFDM a new area of development for Qualcomm?
Davidson: If you go back to the beginning of Qualcomm, OFDM was considered a path instead of CDMA. The company ended up going down the CDMA route because CDMA was better able to handle all the things you want to do on a wide-area wireless network. We believe that to this day.

IDGNS: Are you planning any more acquisitions of companies with OFDM technology?
Davidson: In the last couple of years, our acquisition activity has stepped up. Flarion was clearly the largest deal of the last few years.

IDGNS: Do intellectual property rights play a big role in your acquisition strategy?
Davidson: They can and, clearly in the case of Flarion, there was a double benefit. First and foremost, we got the only team -- to this day -- to deploy a working mobile OFDM system. We also got the intellectual property rights that came along with the business. Our acquisitions are focused on accelerating time to market on a build-versus-buy decision and augmenting engineering resources more than we're out trying to grab patents.

IDGNS: What's driving all the interest in OFDM?
Davidson: We're seeing interest in OFDM because spectrum is becoming available in the 10MHz blocks and wider. From an efficiency standpoint, there's not really a benefit for OFDM over CDMA. But when you get into wider branches of spectrum, it can be a little less complex to implement.

IDGNS: But isn't 4G -- in which OFDM will play a big role -- all about newer, faster services?
Davidson: I think OFDM is really just a spectrum play. And frankly, we don't subscribe to the "4G" term. The applications that I've heard discussed aren't a whole lot different from what is being enabled over 3G today.

IDGNS: Isn't 4G supposed to be a lot faster than 3G?
Davidson: Many are talking about data rates that we don't even get on landline systems today. Yes, you can enable HDTV over these enormously wide pieces of spectrum. But what is the practical cost to the end-user?

IDGNS: So do we really need 4G?
Davidson: There is an existing roadmap within existing 3G technologies that provide the very same and, in most cases, better performance than some of the new technologies being proposed by other groups.

IDGNS: So WiMax and LTE aren't necessary?
Davidson: I look at LTE and UMB as being comparable; WiMax is not comparable to those technologies in terms of performance. There is a mistake in the premise that whatever comes along -- what people are calling 4G -- will be something that supplants the existing networks. We've been saying for several years that it will be about multiple airlinks existing in the market and making them work effectively together.

IDGNS: Let me come back to WiMax: Why isn't it comparable to LTE?
Davidson: Because its original legacy is borne out of the fixed environment, there are immediate engineering trade-offs and performance issues that you come up against. There is this concept of link budget, or how effective a technology is over the airlink. WiMax suffers from poor spectral efficiency because of its heritage as not being a mobile standard.

IDGNS: Do you see any intellectual property rights issues with 4G?
Davidson: We believe that our OFDM, OFDMA, and MIMO portfolio is among the strongest out there and clearly believe that it's applicable to any OFDM/OFDMA systems. Unfortunately, those who are supporting WiMax are trying to make it sound that the IP (intellectual property) picture with this technology is very clear and that it's going to be simple. The IP picture in 3G is much clearer today than what exists in WiMax. The number of companies claiming IP that can be contributed to WiMax is enormous.

IDGNS: Will Qualcomm be active in WiMax in any way?
Davidson: As we said several years back when many were trying to say that Wi-Fi would come and kill 3G, to the extent that our customers want the integration of Wi-Fi into our chipsets, we'll accommodate that. We've said the same about WiMax. We're being pragmatic and view that it will be in the market.

IDGNS:
Nokia CEO Olli-Pekka Kallasvuo said at the company's recent shareholders' meeting that the Finnish manufacturer can't give one company, Qualcomm, a chance to dictate rules for the whole industry. He said the issue is not Qualcomm versus Nokia but rather it's more about Qualcomm versus the rest of the industry. And your opinion?
Davidson: It's amusing to me that Nokia seems to think it's holding up the banner for the entire industry. If not for Qualcomm, there would be far fewer handset manufacturers for them to deal with as competitors and potential competitors. Our business model gives consumers a lot more choice so that Nokia can't dictate pricing into the market. Because we hold intellectual property, Nokia wants to paint us controlling the industry. We enable a lot of competition that causes them a lot of concern -- hence why we're being attacked by them.

The last point is amusing and i tend to agree with Qualcomm on this. Nokia has been dominating the market for long time and its about time other players get in the game.