File Name: planning and optimisation of 3g and 4g wireless networks .zip
Carl Friedrich Gauss 7, Castelldefels, Spain. Planning future mobile networks entails multiple challenges due to the high complexity of the network to be managed. Beyond 4G and 5G networks are expected to be characterized by a high densification of nodes and heterogeneity of layers, applications, and Radio Access Technologies RAT.
In this context, a network planning tool capable of dealing with this complexity is highly convenient. The objective is to exploit the information produced by and already available in the network to properly deploy, configure, and optimise network nodes. This work presents such a smart network planning tool that exploits Machine Learning ML techniques.
The proposed approach is able to predict the Quality of Service QoS experienced by the users based on the measurement history of the network.
We select Physical Resource Block PRB per Megabit Mb as our main QoS indicator to optimise, since minimizing this metric allows offering the same service to users by consuming less resources, so, being more cost-effective. Two cases of study are considered in order to evaluate the performance of the proposed scheme, one to smartly plan the small cell deployment in a dense indoor scenario and a second one to timely face a detected fault in a macrocell network.
Nowadays, we are assisting to the definition of what 5G networks will look like. The main vendors are publishing numerous white papers presenting their view on 5G networks and architectures. What is clear from all these converging visions is that network management in beyond 4G and future 5G networks has to face a whole new set of challenges, due to ultradense deployments, heterogeneous nodes, networks, applications, and Radio Access Networks RANs and also heterogeneous spectrum access through novel technologies, such as Long Term Evolution Unlicensed LTE-U and License Assisted Access LAA , all coexisting in the same setting, the need to manage very dynamic networks where part of the nodes is controlled directly by the users e.
In this context, it is already widely recognized that the network needs to establish new procedures to become more intelligent, self-aware, and self-adaptive. However, this vision needs to be further developed in 5G considering the huge complexity of these networks.
The main objective of network management is then to make the network more self-aware, by exploiting and analysing data already generated by the network this is expected to drive network management from reactive to predictive and more self-adaptive by exploiting intelligent control decisions tools, offered by ML, based on learning and experience. In this paper, among many network management problems, we focus on smart network planning, which gives particular emphasis to the QoS offered to the users and the resources used by the operator to offer it.
We believe that the use of smart network planning tools is crucial and inevitable for operators running multi-RAT, multivendor, multilayer networks, where an overwhelming number of parameters needs to be configured to optimise the network performance. From an industry perspective, the market of commercial planning tools aims at providing a complete set of solutions to design and analyse networks [ 5 — 8 ].
For instance, in [ 9 ], the authors present an open-source network planning tool that includes different planning algorithms to analyse the network under different failures and energy efficiency schemes. However, these works in general focus on several configuration scenarios, RF coverage planning, network recovery test, traffic load analysis, and forecasting traffic, among others, and not directly on QoS offered to end-users and the resources that mobile network operators need to offer.
Other works are more targeted to QoS estimation, but not in the area of network planning. The literature already offers different works targeting the problem of QoS prediction and verification, such as [ 10 , 11 ]. In our preliminary work [ 12 ], we focus on complex multilayer heterogeneous networks where we predict QoS independently of the physical location of the UE, that is, based on data learned throughout the whole network.
Preliminary results show that by abstracting from the physical position of the measurements we can provide high accuracies in the estimations of QoS in other arbitrary regions. Furthermore, the results presented in [ 13 ] show that data analysis achieves better performance in a reduced space rather than in the original one. This paper presents a smart network planning tool that works in two steps. First, it estimates the QoS at every point of the network based on user-collected measurements, which may have been taken at different time instants and anywhere in the heterogeneous network.
We perform this estimation through supervised learning tools. This results in an appropriately tuned QoS prediction model, which is then integrated in the next step. Second, we adjust the network parameters in order to reach certain network objectives by evaluating different combinations and calculating the resulting QoS based on the model of step 1.
Network objectives are set in terms of PRB per Mb. We focus on this specific QoS indicator because it combines information that is relevant to the operator PRBs and other that is relevant to the end-user Mb of data into a single metric.
Therefore, the minimization of this indicator allows serving users with an improved spectral efficiency and offered QoS. More specifically, to carry out this optimisation we take advantage of GAs, which are stochastic search algorithms useful to implement learning and optimisation tasks [ 14 ], and so are adequate for our purpose. In order to evaluate the performance of the proposed scheme, we consider two use cases. The first one is in a densified indoor scenario, and the second one is in a more traditional macrocell scenario.
In the first use case, we focus on how to plan a dense small cell deployment inspired in a typical 3GPP dual stripe scenario [ 15 ], where the parameters to adjust are the number of small cells and their position. In the second use case, we deal with self-healing aspects in macrocellular scenarios. We focus on readjusting the parameters of the surrounding cells to quickly solve an outage problem by automatically adjusting their antenna tilt parameters. We evaluate the performance of the proposed planning tool through a network simulation campaign carried out over the 3GPP compliant, full protocol stack ns-3 LENA module.
We show that the proposed ML-based network planning, differently from other closed optimisation approaches, is extremely flexible in terms of application problem and scenario, and, in this sense, it is generic.
Without loss of generality, particular attention has been focused on how to plan a dense 4G small cell deployment and how to quickly solve an outage problem. As can be seen, the approach is equally valid and shows good performance in both studied use cases, characterized by very different optimisation problems and scenarios.
As for the RAT, we have focused on 4G and its evolution. The same technique can be applied also to 2G and 3G technologies. This paper is organised as follows. The general approach is described in Section 2. It describes the ML-based network planning tool, its main design principles, and algorithms we use to build it. In Section 3 , we discuss the specific design details and tuning of the QoS prediction model and we put all the pieces i.
In Section 4 , we present the details of the two use cases to which the tool is applied, the simulation platform, and the simulations results. Finally, Section 5 concludes the paper.
We propose designing a network planning tool, which works in two steps. First, we propose to model the QoS through the analysis of data extracted from the networks in the form of measurements. This phase requires to first prepare the data and then to analyse it. And second, we keep on adjusting the parameters and analysing the impact on QoS based on the previous model. In this way, the performance of the network is optimised to meet certain operator targets.
Figure 1 presents the different phases required by the proposed network planning. This process aims at transforming data into a meaningful format for the estimation at hand.
The target is to integrate and prepare large volumes of data over the network to provide a unified information base for analysis. To do this, we follow the Extract-Transform-Load ETL process, which is responsible for pulling data out of the source and placing it into a database.
It involves 3 main steps: data extraction E , whose objective is to collect the data from different sources; data transformation T , which prepares the data for the purpose of querying and analysis; data loading L , which loads the data into the main target, most of the cases into a flat file. This process plays an important role for the design and implementation of planning for future mobile networks.
The objective is to create a data structure that is able to provide meaningful insights. Some examples of the kind of sources available in mobile networks are shown in Table 1 [ 3 ]. Here the data are classified based on the purpose for which they are generated in the network. The usage that is given nowadays in the network is also suggested in the last column.
For the purpose of network planning, we plan to extract data reported by the UEs to the network in the form of UE measurements, in terms of received power, received quality, and offered QoS.
Once the data has been collected, we prepare the data for storing, using the proper structure for the querying. The objective of this process is to discover patterns in data that can lead to predictions about the future. We do this by applying ML techniques. The objective of this process is to find the configuration parameters for the optimised network planning based on the information extracted from the previous data analysis process. In the complex cellular context, we need to deal with several network characteristics that introduce high complexity, for example, the very large number of parameters, the strong cross-tier interference, fast fading, shadowing, and mobility of users.
In order to deal with these issues and to guarantee an appropriate network planning, in this work, we propose to use GAs, which allow avoiding some of the problems of typical closed optimisation techniques e. More specifically, they work with chromosomes i. For each of the chromosomes, they calculate its fitness score based on a given objective function in our case, the QoS predicted by the model [ 17 ].
Then, they select the chromosomes with the best fitness score and generate better child chromosomes by combining the selected ones and they keep on iterating until the objective function of the chromosomes generated reaches the performance target. In this way, GAs perform parallel search from a population of points, which represent the values of the different parameters to be tuned in the scenario and, jointly with other techniques, they have the ability to avoid local minima and use probabilistic search rules.
We estimate the QoS at every point of the network based on measurements collected in different moments in time and from other regions of the heterogeneous network, that is, based on the measurement history of the network. To do this, we consider the data preparation and data analysis processes of the network planning tool. As mentioned previously, the objective of these 2 processes is to extract, prepare, and analyse the information already available in the network to provide insightful information from the analysis of it.
In fact, in this kind of estimations, ML techniques can be very effective to make predictions based on observations. We propose using SL, since among many applications it offers tools for estimation and prediction of behaviours. In particular, we focus on a regression problem, since we want to analyse the relationship between a continuous variable PRB per Mb and the data extracted from the network in the form of UE measurements.
Many regression techniques have been developed in the SL literature, and criteria to select the most appropriate method include aspects such as the kind of relation that exists between the input and the output or between the considered features, the complexity, the dimension of the dataset, the ability to separate the information from the noise, the training speed, the prediction speed, the accuracy in the prediction, and so on.
We focus on regression models, and we select the most representative approaches. We then use ensemble methods to sub-sample the training samples, prioritizing criteria such as the low complexity and the high accuracy. We then build a dataset of user measurements, based on the same data contained in the Minimization of Drive Tests MDT database.
The dataset contains training samples rows and features columns and is divided into sets, the training set to train the model and the test set to make sure that the predictions are correct.
That training data develop a predictive model and evaluate the accuracy of the prediction, by inferring a function , returning the predicted output. The input space is represented by an -dimensional input vector. Each dimension is an input variable. In addition, a training set involves training samples. Each sample consists of an input vector and a corresponding output of one data point. Hence, is the value of the input variable in training sample , and the error is usually computed via or with the root mean square error.
In addition to regression analysis, we exploit Unsupervised Learning UL techniques for dimensionality reduction to filter the information in the data that is actually of interest thus reducing the computational complexity while maintaining the prediction accuracy. And after that, the SL techniques under evaluation are then applied. Details for each step are given in the following, and the whole process is depicted in Figure 2.
The data we take into account comes from mobile networks, which generate data in the form of network measurements, control, and management information Table 1. As we mentioned previously, we focus on MDT functionality, which enables operators to collect User Equipment UE measurements together with location information, if available, to be used for network management, while reducing operational costs.
Author: J. An overwhelming development has taken place in voice and data communication over the las twenty years as the industry evolved from fixed to mobile and wireless communication. This development is supported with new technologies and evolving networks from the first generation 1G , 2G, 3G and the fourth generation 4G mobile wireless communications. During this evolution and revolution in telecommunications, the industry also changed from circuit switched networks to packet switched networks in 3G and 3G. Hence the planning of telecommunication networks has equally changed significantly. By providing the necessary background and technical content to understand stay abreast of how to plan the new network types, Planning and Optimisation of 3G and 4G Wireless Networks explores the idiosyncrasies of how to plan the various types of wireless networks. Packed with details of the technologies that support each network type, this cutting-edge reference leads the reader step by step on how to plan and optimize various types of wireless networks.
Huawei ConfidentialPage 8 Preparations for RF OptimizationChecklist Network plan, network structure diagram, site distribution, site information, and engineering parameters Drive test results such as service drop points and handover failure points in the current area Reference signal received power RSRP coverage diagram Signal to interference plus noise ratio SINR distribution diagram. This knowledge transfer is obtained through hands-on experience using UE based diagnostic tools and scanner tools. Understanding of system topology and air interface parameters of LTE. Based on these detection results, related parameters are adaptively configured and adjusted to effectively accelerate 5G deployment. Good knowledge of Huawei equipment.
An overwhelming development has taken place in voice and data communication over the las twenty years as the industry evolved from fixed to mobile and.
A sophisticated in-building RF planning tool normally takes a building as an input and designs network systems in the building. Tell us what is your experience with RF technology? Did you work in the field of RF engineering? How was your experience in RF design and implementation?
Jetzt bewerten Jetzt bewerten. A highly practical guide rooted in theory to include the necessarybackground for taking the reader through the planning,implementation and management stages for each type of cellularnetwork. They even contain features of thetechnologies that will lead us to the fourth generation networks. Designing and optimising these complex networks requires muchdeeper understanding.
Register Login. A practical network will have cells of no geometric shapes. RF planning and site design. However, in the CBRS band the spectrum is shared with incumbents and higher tier users in three dimensions—spatial, temporal, and spectral. Note the latitude and longitude from planning tool.
ASSET Radio can be customized through a range of productivity packs, enabling customers to add additional functionality to meet their needs. Newly adopted technologies, such as 5G, need to be deployed on top of existing networks. Our 5G NR modeling includes and supports advanced propagation models, complex antenna arrays and full multi-technology 3D coverage and capacity simulations.
Export RIS format. Closed Access. Filename Description Size OK. Full metadata record.
Skip to search form Skip to main content You are currently offline. Some features of the site may not work correctly. Agbinya Published Engineering.
It includes coverage of the latest wireless network types, planning and optimization methods in the form of: 3G; HSPA and Beyond 3G; WiMAX (fixed and mobile).
Я сейчас же отправлю ее домой. - Боюсь, вы опоздали, - внушительно заявил Беккер и прошелся по номеру. - У меня к вам предложение. - Ein Vorschlag? - У немца перехватило дыхание. - Предложение.
- Неужели фильтры безопасности что-то пропустили. В целях безопасности каждый файл, загруженный в ТРАНСТЕКСТ, должен был пройти через устройство, именуемое Сквозь строй, - серию мощных межсетевых шлюзов, пакетных фильтров и антивирусных программ, которые проверяли вводимые файлы на предмет компьютерных вирусов и потенциально опасных подпрограмм. Файлы, содержащие программы, незнакомые устройству, немедленно отвергались.
Она все еще не могла свыкнуться с мыслью о шифре, не поддающемся взлому.
The task of optimizing service quality in wireless networks is a continuous research that requires the design of efficient channel allocation schemes.SГіfocles O. 14.05.2021 at 03:11
Carl Friedrich Gauss 7, Castelldefels, Spain.Zoraida M. 14.05.2021 at 19:17
the rapid developments of wireless communications tech-. nologies. Wireless networks such as 3G/4G and WLAN/WMAN. play an important.Emerenciana O. 15.05.2021 at 00:14
Corbett classical rhetoric for the modern student pdf cell by stephen king free pdfSimplicio R. 18.05.2021 at 23:03
Nolo guide to social security disability pdf corbett classical rhetoric for the modern student pdf