 |
|||||||||||||||||||
Solutions to Improve 3G Networks using Multi-User Detection and Channel Estimation
By Daniel massicotte, Founder President and Chief Technology Officer, Axiocom inc. Messaoud Ahmed-Ouameur, Director R&D, Axiocom inc.
1. The receiver - a critical component of any base station. The receiver main functions are to faithfully reproduce the transmitted information and take it to the appropriate address. However, because of the multiple-path phenomenon (multiple reflections, diffractions and attenuations of the signal transmitted through the communication channel), the receiver must also “manage” the presence of several types of interference, namely multiple-access interference (MAI) and intersymbol interference (ISI)
These two types of interference are the main problems of the 3G digital signal transmission. Their presence seriously inhibit the operations of the network, because it limits base station coverage, user capacity (number of simultaneously processed users by the base station), quality of service, useful lifetime of the base stations and cellular batteries. It also distorts the quality of transmitted data and delays the deployment of 3G networks for high-speed data transmission.
Consequently, in order to achieve adequate performance levels, the base station receiver must necessarily be able to efficiently eliminate MAI and ISI interference. Otherwise, interference impairment to the network would mean problems of non-accessibility to the network, disconnection and loss of information for users. As for network operators, it would represent waste of billed airtime, loss of subscribers and loss of revenues. Currently, several digital signal-processing techniques propose hypothetical solutions that eliminate, or at least offset the effects of said interference. Because they are hypothetical, few of these techniques take into consideration the microelectronic specifications requirements to their implementation, although imperative to operational, economic and commercial prerogatives of the industry. Worse, even fewer techniques succeed at being potentially applicable at the commercial level. The Correlator-Rake receiver, representing the standard in terms of digital signal correction for CDMA in second generation (2G) cellular networks, is also used by third generation (3G) cellular network manufacturers, even though it cannot cancel efficiently multiple-user interference. The communication channel model is directly responsible for this constraint, because the Correlator-Rake receiver algorithm does not provide for the identification and processing of interference due to multiple subscribers using the same frequency. Considering that, as it happens, 3G transmission technologies favour simultaneous sharing of the same frequency by multiple users, the Correlator-Rake receiver could therefore hardly meet the performance levels required, and demanded by this type of network and by the industry.
2. Axiocom solutions – at the Digital Signal Reception LevelIn order to meet the needs of digital communication markets in the field of 3G cellular telephony, Axiocom has made a breakthrough that is significant and widely recognized by the industry, by developing high-performance interference elimination methods for all 3G transmission technologies for WCDMA (Wideband Code Division Multiple Access), CDMA2000 and TD-SCDMA. As importantly, not only the solutions developed by Axiocom prove to be superior to current methods, but they also meet the telecommunications industry's commercial requirements in terms of operations with quick and reliable call connectivity and processing, functionality by increasing voice and data quality and faithful reception of transmitted information and profitability by a superior return on investment thanks to a technology that proposes a greater call processing capacity at a lower cost (the integration costs of the Axiocom technology to 3G base stations are significantly lower than other proposed technologies). When integrated to the base station receiver (uplink), the Axiocom solution includes two components, namely the ‘Axcest' channel estimator that provides channel components to the interference elimination filter – the detector (e.g. Rake and MUD) and the ‘Axnet' multi-user detector (MUD) that evaluates the data transmitted by mobile stations. All Axiocom products can be applied to various types of 3G base stations, regardless of transmission technologies (WCDMA, CDMA2000, TD-SCDMA or HSDPA).
2.1 Performance for WCDMA 3G base stationsAxiocom‘s research team has developed and compared its multi-user detection and channel estimator solutions for the WCDMA transmission technology with the conventional Correlator-Rake equalization method. Following is an example of the range capacity of the Axiocom's solution vs. the current industry standard, the Correlator combined to the Rake long code receiver.
A comparison was carried out with the solutions most often mentioned by competition, the hard multistage parallel interference cancellor (Hard-MPIC) and the soft multistage parallel interference cancellor (Soft‑MPIC). Moreover, this study was done in presence of better scenarios context for the MPIC structure by including a decision feedback (called in the figures DF-Hard-MPIC and DF‑Soft-MPIC) and a pilot cancellation process. It is known that the decision feedback technique is essential for the MPIC structure to perform but increase the implementation complexity to parallelize the processing (ex: pipeline) on a device due to the data dependency problem. Compared with the Rake long code receiver using the conventional Correlator channel estimator for taps estimation assuming perfect knowledge of the delays, the Axiocom multi-user detection receiver (Axnet) has, in spite of its low-level algorithmic complexity, generated performances that are superior to those of the Correlator-Rake and DF-Hard-MPIC receivers in terms of coverage (transmission power of the mobile station) and capacity for processing a number of simultaneous users.
§ Results show equivalent performance of Axnet compared to the high implementation complexity candidate DF-Soft-MPIC.
§ When compared to the current technology Rake-long Code, and the candidate DF-Hard-MPIC, with a given raw bit error rate (BER) of 5% and a transmitting power level to achieve a signal to noise ratio (SNR) of 8 dB, Axiocom solution (Axnet) provides future WCDMA network operators to simultaneously process up to 3 times more users over currently WCDMA network operators using the Rake-long code.
§ When using the Axnet solution, at a level of SNR of 8 dB, WCDMA base stations receivers can process simultaneously more than 7 users, while WCDMA base stations receivers using the Rake-long code and DF-Hard-MPIC can process simultaneously no more than 2 and 4 users respectively.
§ Axiocom's solution for WCDMA base stations allows network operators to either lower their transmitting power or increase the distance between their transmission sites (coverage).
2.3 Channel estimation to boost the Rake in WCDMA 3G base stations
§ Given a raw BER of 10% and a transmitting power level to achieve an SNR of 8 dB, Axiocom's solution was able to provide network access to an additional 10 users (50% more users) when compared to the Correlator-based channel estimator using one antenna.
§ Using Axiocom's channel estimator (Axcest), WCDMA base stations receivers can process simultaneously 30 users at a level SNR less than 8 dB, while WCDMA base stations receivers using the Correlator-based channel estimator require an SNR of at least 12 dB for the same number of users.
§ With 4 dB improvement in performance over the Correlator-based channel estimator when using Rake-long code receivers, Axiocom's solution allows a reduction of transmitted power of over 150%.
§ Axiocom's solution for WCDMA base stations allows network operators to either lower their transmitting power or increase the distance between their transmission sites.
§ The low algorithmic complexity of Axiocom's solution equivalent to the Correlator-based channel estimator allows its integration into existing BTS technologies in the same material.
4. Implementation complexity – a key component The low implementation complexity level is the key of Axiocom's interference cancellation solution “Axnet”, allowing a faster integration into 3G base station technologies, at a lower cost than the competing solutions. Compared to the multiple-access interference elimination technique used in the actual 3G networks “Rake”, Axnet solution shows only a complexity level upto 5, while Soft‑MPIC shows a complexity level up to 17 times higher (with or without decision feedback applied to the MPIC methods). Not only Axiocom's technology can reach more performances (dB gain), but in addition, Axnet can have a complexity level 4 to 6 times lower than the Soft-MPIC. Moreover, compared to other solutions, our method has the advantage of reducing the complexity according to the transmission speed.
The results analysis confirms that only Axiocom's solution can reach 100% BTS users on a technology platform with les than 5 times the actual complexity level based on the Rake. We have demonstrated the FPGA implementation based Axiocom's solution of all users on the family devices of Virtex II and Virtex II PRO of Xilinx.
§ To carry out high rate data transmission reliably, Axnet offers a superior performance to one of the most competing technologies based on DF-Soft-MPIC while maintaining a substantial complexity reduction by a factor of 4 to 6 at full capacity.
§ Unlike DF-Hard MPIC and DF-Soft MPIC, Axnet exhibits a diminishing complexity as data rates increase. § The low algorithmic complexity of Axiocom's solution allows its integration into existing BTS technologies.
§ Axiocom's solution for WCDMA base stations allows manufacturers to include Multi-Users Detector solutions with a minimum of cost.
Beyond the arithmetic complexityAnother important aspect to compare the implementation complexity is the algorithmic structure such as regularity, recursiveness, data flow, memory quantity and inherent parallelism – all qualities intrinsic to the Axiocom's solution. In this study, we have not included these aspects to compare MUD methods. However, an obvious consideration can be observed with the decision feedback structure of MUD. Indeed, even if the decision feedback structure might have relatively the same complexity level, the main drawback is the lack of parallelism that can be exploited, especially for the MPIC caused by data dependencies. In fact, a DF-MPIC at instant n and for user k needs to wait for all users so that the kth user proceeds to detect the current data before processing its own data. Such a structure looses its parallelism to apply pipeline or parallel techniques and to become serial operation limited for sequential DSP implementation. Hence, the DF-MPIC will always be limited by the DSP clock speed to respect the computational time imposed by the 3GPP time frame. Noted that, Axiocom's solution do not use decision feedback structure to exploit the parallel implementation techniques.
5. Multi-users detection and channel estimation benefits
Translated into operational and economic terms, a SNR gain of 4 dB corresponds to about 60% increase in geographical coverage per base station and a cost reduction for network operators upto $345 million US when implementing a 3G network of 5,000 base stations.
The performance level achieved by the Axiocom Axnet multiple user detection receiver, and its Axcest channel estimator offers network operators several benefits, namely :
6. ConclusionAxiocom‘s research team has developed and compared its multi-user detection (MUD) and channel estimator solutions for the 3G cellular network with the conventional Correlator-Rake receiver and Multistage Parallel Interference Cancellation (MPIC) techniques. From the implementation complexity point of view, a fair implementation complexity analysis is addressed. Axnet shows better performances and lower implementation complexity when compared to different versions of the PIC in presence of the Correlator based channel estimation and Axcest shows substantial average gains of 4 dB's with an equivalent implementation complexity into a general purpose DSP implementation. Axiocom's solutions respond to various types of 3G base stations, regardless of transmission technologies (WCDMA, CDMA2000, TD-SCDMA or HSDPA) and a solution for HSPA Evolution.
Authors
For more information: Daniel Massicotte, Ph.D. President Founder and Chief Technology Officer 1350, Royale Street, suite 401 Trois-Rivières (Québec) CANADA G9A 4J4 Email: daniel.massicotte@axiocom.com Phone : 819.697.2946 Fax : 819.697.0917
|
|||||||||||||||||||
|
|||||||||||||||||||
