AUTOMATED CONTROL SYSTEMS
HETEROGENEOUS COMMUNICATION NETWORKS IN NETWORK SYSTEMS
MONITORING
Olimpiev A.A.,
JSC Scientific Research
Institute "Rubin",
Sherstyuk Yu.M., Doctor of Technical Sciences, Associate Professor, JSC "Research Institute" Rubin ", [email protected]
Keywords:
information system, object-oriented approach, learning automata, finite automata, grammars.
ANNOTATION
General trends in the development of domestic automated communication control systems and existing technologies for creating systems of this class are considered. The disadvantages of traditional approaches to the creation of information models, which are the basis for the construction of information systems and consist in the excessive growth of the content and structure of the model for the presentation and storage of information, are highlighted.
A formal model of the object representation of a heterogeneous communication network is proposed, which makes it possible to quickly calculate the integral state of a communication network and its elements. A communication network is represented as a group of objects interacting by means of message transmission. Each object is an instance of some class and is represented as a state machine with arbitrarily complex behavior. The content of the model does not depend on the data transmission technologies used in the network and the composition of the equipment, which makes it capable of adapting to evolutionary changes in the communication network.
As a method for optimizing the collection of monitoring data designed to update the state of the object model, an approach based on a system of learning automata is chosen. This approach allows you to achieve high efficiency in updating the state of the object model in the absence of information about the network infrastructure by adapting to the response time of the system.
AUTOMATED CONTROL SYSTEMS
Introduction
Within the framework of the creation of automated communication control systems (ACS) at the operational and technical level, one of the most urgent tasks to be solved is the problem of adequate information display of a controlled communication network as an object of monitoring and control (WMD). The information model serving as a component of the decision-making support system in the control loop should reflect the composition, connections and characteristics of the elements of WMD that are most consistent with the current state of the WMD and its components.
Currently, there are approaches to representing networks of a similar nature (see, for example,), but the dimension of the resulting representations is extremely high. In addition, static accounting of all network elements is not advisable - it is extremely resource-intensive, and will duplicate the data that can be obtained from technological monitoring tools.
An obstacle to the creation of an adequate model of a communication network is the existing inconsistency of the conceptual and information models of the operational-technical and technological levels of management, which consists in the fact that at the technological level, the elements of the communication network are represented by their own management information bases (Management Information Block - MIB), which take into account their specifics in terms of software and / or hardware implementation, and on the operational and technical features of the communication network elements must be "hidden" from the user - the network layer assumes the operation of concepts that are common for equipment of the same type with different MIBs.
Considering that the technological level when creating an ACS is objectively specified and unchanged, the problem generated by this contradiction cannot be solved within the framework of the "accounting" information model - it must be supplemented with some kind of computational formalism, which can be an object representation model.
Formal Object Model
communication network views
The essence of the computational formalism of the object representation of modern communication networks can be defined as follows:
1). The central concept of the model is the concept of an object - an abstract entity characterized by its parameters and behavior:
o =
The object state parameter can take a value from a fixed set - ("norm", "accident", "warning", ...).
2). The following relationships exist on a set of objects:
"the whole is a part of the whole" (Risa);
supplier-consumer (Ruse);
"interaction" (Rcon). Sd = (O, Risa, Rcon, Ruse),
where Sd is a mapping from a set of relations to a set of objects.
3). Each object is an instance of a certain class. Classes form a hierarchy with the ability to inherit parameters and methods.
V o e O 3 cl e CL: o => cl, where CL is the set of all classes.
4). In essence, classes and their corresponding objects can be conditionally divided into three groups:
"terminators" - nodes of graphical representation of communication networks;
"connectors" - edges of the graph representation of communication networks;
"aggregators" - abstract entities - a logical combination of objects into a group with the ability to calculate its integral state.
5). Many of the object's methods have a mapping to input messages.
Input messages include: creation / deletion of an object; creating / deleting object relations; changing the state of interacting objects; changing the values of the object's parameters (including the functioning parameters calculated from the monitoring data).
6). An object is considered as a finite state machine capable of receiving messages and, based on them, change its state and / or generate messages. The rules for navigating and generating messages can be as complex as you like.
The operation of the machine can be written as follows:
st (tm) = v (x, st (ti)), (y) = ф (x, st (ti)), where st is the state of the automaton; x - input messages, y - output messages; x, y with S, where S is the set of all possible messages.
7). The "object manager" acts as a component of supporting the computing environment, which performs the following actions:
creating and deleting objects;
analysis of incoming messages and their transmission to recipient objects;
generating messages for creating / deleting relationships over objects;
formation of messages taking into account relations over objects.
The "Object Manager" can be thought of as a push-down machine:
^ o (Q cho, GM, Gvx, Gvykh, G, Ib),
where Гвх, Гвх с S are the grammars of the input and output tape, respectively; ГМ = Г] с Г2, Г] с S, Г2 = (
The mapping G: Q x Гм х Гвх ^ Q х Гм х Гвх, defines the set of rules for transitions between states.
eight). The incoming messages to the "object manager"
HIGH TECH IN EARTH SPACE RESEARCH
AUTOMATED CONTROL SYSTEMS
hype can be generated as a reaction to one of the following events:
changing the state of an object;
change of credentials;
detection of significant events at the level of network elements.
nine). The actualization of the network state is carried out by the gateway of interaction of the means of the technological and operational-technical levels on the basis of a set of significant events occurring in the monitoring environment.
The set of significant events over a period of time D / at the level of network elements can be represented as follows:
Y (D) = DVshv (N) and UA (D), where DBshv is the dynamics of the parameters M1V, UA (D /) is a set of external influences on network elements, D1 = / k- / k-1 is the time interval between polls of funds technological monitoring.
DVShv (D0 = 1ДП „№, where m =, N = is the set of all network elements, r =, / p is the number of equipment classes, ng is the number of instances of this class.
D = exp.H, j =)), 1 (]) = ^] (mxn (M]), f, y, U),
where tsh (D /) is the minimum allowable polling time of the> th network element, f is the frequency of polling the network element by means of technological monitoring, Yj is the number of external influences on the j-th network element.
It is advisable to solve the problem of optimization D / by using learning automata, the work of which can be represented as:
АМ = (ШС, 2, X, Zo, DO), where Ш = (^ 1, m2, ... mn) is the memory vector, C is the penalty matrix, 2 is the random control operator, X is the control vector, X = 2 (X-ъ DX), X =<Д/, П>, O "= F (Pshv), - conditions assigned by the upper-level system or operator, DO, = D ^ b DD (Xr-1), 2o).
Based on Y (D /), the gateway generates a set of messages that arrive on the input tape of the object manager.
Conclusion
Due to the presence of the mechanisms described above, the object model can be figuratively considered as a kind of neural network, in which an external stimulus (credentials, monitoring data) leads to the creation / removal of objects and / or the execution of a fading process of neuronal excitation, propagating along the information model of the network - the process of state updating information model.
An important result of using the described mechanisms is the ability to quickly obtain information about the state of not only an individual piece of equipment or a communication line, but also an integral assessment of the state of the communication network as a whole.
Literature
1. Grebeshkov, A. Yu. Standards and technologies of communication network management [Text]: Manuscript. - M .: Eco-Trends, 2003 .-- 288 p.
2. Sherstyuk, Y. M. Architecture of technological management of telecommunications [Text] / Y. M. Sherstyuk,
B. D. Zaripov, M. D. Rozhnov, I. L. Saveliev // Telecommunication technologies. - 2006. - Issue. 2.S. 33-40.
3. Sherstyuk, Yu. M. Architecture and main directions of development of an automated control system of a unified information and telecommunication system [Text] // Telecommunication technologies. - 2007. - Issue. 3.
4. Olimpiev A. A. Unification of the representation of communication networks based on the object approach [Text] / A. A. Olimpiev, M. D. Rozhnov, Yu. M. Sherstyuk // V St. Petersburg interregional conference “Information security of Russian regions - 2007 (IBRR-2007) ", St. Petersburg, 2325 October 2007: Proceedings of the conference. Section: Information security of telecommunication networks. - SPb .: SPOISU, 2008.S. 60-66.
5. Sherstyuk Yu.M. Proposal for solving the problem of updating the state of a heterogeneous telecommunication network [Text] / Yu. M. Sherstyuk, AA Olimpiev // Questions of radio electronics. Ser. SOIU. - 2012. - Issue. 2.S. 5-10.
HETEROGENEOUS COMMUNICATION NETWORKS IN THE NETWORK MONITORING SYSTEM
JSC "Rubin" Research Institute, [email protected]
Sherstyuk Y., Doc.Tech.Sci., Docent, JSC "Rubin" Research Institute, [email protected]
In the article are some general trends in the development of network management systems. Considered the traditional approach to the creation of such systems.
A formal model of object representation of a heterogeneous network, which allows quickly calculate the integral state of the communication network and its elements. The communication network is represented as a band machine interacting via messaging.
As an optimization method of data collection for monitoring, intended to update the state of the model is chosen approach, which is based on a system of learning automata. This approach allows us to achieve high efficiency of updating the state of the information model in the
absence of information about the network infrastructure. Keywords: information system, an object-oriented approach, learning automata, finite automata, grammars.
1. Grebeshkov, A 2003, "Standards and technologies of control of communication networks", Moscow, 288 pages.
2. Sherstyuk, Yu 2006, "Architecture of means of technological telecommunication management", Telecommunication technologies, vol. 2, pp. 33-40.
3. Sherstyuk, Yu 2007, "Architecture and main directions of development of an automated control system of uniform information telecommunication system", Telecommunication technologies, vol. 3, pp. 5-14.
4. Olimpiyev, A 2008, "Unification of representation of communication networks on the basis of object approach", the V St. Petersburg interregional conference "Information Security of Regions of Russia-2007 (IBRR-2007), pp. 60-66.
5. Sherstyuk, Yu 2012, "Proposal according to the solution of the task of updating of a status of a heterogeneous telecommunication network", Radiotronics Questions, vol. 2, pp. 5-10.
From Wikipedia, the free encyclopedia
| This article is being proposed for deletion. You can find an explanation of the reasons and the corresponding discussion on the page Wikipedia: Removed / December 15, 2015. |
Heterogeneous computer network- a computer network that connects personal computers and other devices with different operating systems or data transfer protocols. For example, a local area network (LAN) that connects computers running Microsoft Windows, Linux, and MacOS operating systems is heterogeneous. The term "heterogeneous networks" is also used in wireless computer networks, where various technologies are used to connect. For example, a wireless network that provides access over a wireless local area network and is capable of providing access by switching to cellular is also called a heterogeneous network.
HetNet
Mention of technology HetNet often means using multiple types of access points in a wireless network. The WAN can use macro cells, pico cells and / or femto cells to provide coverage in environments with different types of terrain, from open spaces to office buildings, homes and underground spaces. Cellular experts define HetNet as a network with complex interactions between macro cells, small cells and, in some cases, WiFi network elements, all of which are used together to provide mosaic coverage with handoff capability between network elements. ARCchart research predicts that HetNets will drive the mobile infrastructure market, which is valued at approximately $ 57 billion by 2017.
Semantics of "Heterogeneous computer networks" in telecommunications
From a semantic point of view, it is important to note that the concept heterogeneous networks may have various meanings in the field of wireless telecommunications. For example, it can mean a paradigm of well-integrated and ubiquitous interoperability between different protocols using different coverage areas. HetNet). In other cases, this can mean uneven spatial distribution of users or wireless access points (see. Spatial Inhomogeneity). Therefore, the use of the term “heterogeneous networks” without context can cause confusion in the scientific literature when reviewing the work by others. In fact, confusion may be exacerbated in the future, especially in light of the fact that the HetNet paradigm can also be viewed from a “geometric” perspective.
see also
Write a review on the article "Heterogeneous Computer Network"
Literature
An excerpt characterizing the Heterogeneous Computer Network
Rostov in the campaign allowed himself the liberty to ride not a front-line horse, but a Cossack one. Both an expert and a hunter, he recently got himself a dashing Don, large and kind playful horse, on which no one jumped him. Riding this horse was a delight for Rostov. He thought of the horse, of the morning, of the doctor, and never once thought of the impending danger.Before Rostov, going into business, was afraid; now he did not feel the slightest feeling of fear. Not because he was not afraid that he was used to fire (you cannot get used to danger), but because he learned to control his soul in the face of danger. He was accustomed, going into business, to think about everything, except for what, it seemed, would be more interesting than anything else - about the impending danger. No matter how hard he tried, no matter how much he reproached himself for cowardice during the first period of his service, he could not achieve this; but over the years now it has become of itself. He was now riding next to Ilyin between the birches, occasionally tearing off leaves from branches that came under his arm, sometimes touching the horse's groin with his foot, sometimes giving away the smoked pipe to the hussar riding behind, with such a calm and carefree look, as if he was riding ride. He was sorry to look at the agitated face of Ilyin, who spoke a lot and with anxiety; he knew from experience that agonizing state of expectation of fear and death in which the cornet was, and he knew that nothing but time would help him.
As soon as the sun appeared on a clear strip from under the clouds, the wind died down, as if it did not dare to spoil this lovely summer morning after a thunderstorm; drops were still falling, but already sheer - and everything was quiet. The sun came out altogether, appeared on the horizon and disappeared in a narrow and long cloud that stood above him. A few minutes later, the sun appeared even brighter on the upper edge of the cloud, tearing its edges. Everything shone and shone. And along with this light, as if answering it, the shots of guns rang out ahead.
Before Rostov had time to ponder and determine how far these shots were, the adjutant of Count Osterman Tolstoy galloped from Vitebsk with the order to trot along the road.
The squadron drove around the infantry and the battery, which was also in a hurry to go faster, went downhill and, passing through some empty village without inhabitants, again climbed the mountain. Horses began to lather, people flushed.
- Stop, be equal! - the divisional team was heard ahead.
- Left shoulder forward, step march! - commanded ahead.
And the hussars along the line of troops passed to the left flank of the position and stood behind our uhlans, who were standing in the first line. On the right was our infantry in a thick column — these were reserves; higher up on the mountain were visible in the clean clean air, in the morning, oblique and bright, illumination, on the very horizon, our cannons. Enemy columns and cannons were visible ahead of the ravine. In the hollow we could hear our chain, which had already entered into action and merrily flipped with the enemy.
Rostov, as from the sounds of the most cheerful music, felt cheerful in his soul from these sounds, which had not been heard for a long time. Trap ta ta tap! - clapped, then suddenly, then quickly, one after the other several shots. Again everything fell silent, and again it was as if the crackers were cracking, on which someone was walking.
The hussars stood in one place for about an hour. The cannonade also began. Count Osterman and his retinue rode behind the squadron, stopping, talking to the regiment commander, and drove off to the cannons on the mountain.
Following the departure of Osterman, the lancers heard the command:
- In the column, line up for the attack! - The infantry in front of them doubled platoons to let the cavalry through. The lancers set off, swaying their peak with weathercocks, and trotted downhill towards the French cavalry, which appeared to the left under the mountain.
Network heterogeneity- heterogeneity of communication and hardware configuration, as well as software in structured networks.
Methods:
· Encapsulation
It is used in cases when: - it is necessary to organize the exchange of data between two networks, built using the same technology, for different physical. Wednesdays; -when 2 networks are not connected directly, but through intermediate networks using excellent technologies.
Principles: 1. Transport packages protocols that need to be sent over the transit network are encapsulated; 2. After passing through the transit network, the reverse process of decapsulation and forwarding to the addressee takes place. Advantage: fast and easy to implement method
Flaw: does not provide interaction with the nodes of the transit network.
· Broadcast - coordination of 2 protocols by converting the format of messages coming from one network to the format of another network. Bridges, switches, routers, and gateways can broadcast. Flaw: laborious, with t. processing power of methods that can decrease the speed of data transmission over the network.
· Multiplexing
A method when nodes are simultaneously installing and configuring the simultaneous operation of several protocol stacks at once, which allows them to process messages from nodes of heterogeneous subnets. 
Multiplex. protocols- software that performs the task of determining the isp-th received message of the protocol stack. Advantages : - a simpler method to implement than translation; - overcoming bottlenecks in the network; no queues to a single gateway. Disadvantages: the administration and monitoring of the network performance becomes more complicated; redundancy requires additional resources to the workstation.
21. Routing of the network layer. Route table. Routing algorithms. The concept of a metric.
21. Routing packets. Route table. Routing algorithms. The concept of a metric.
Routing - a mechanism that allows delivery of packets from one node to another in a structured heterogeneous network. Routing can be done:
· on channel level (through bridges and switches).
Restrictions interactions arising at the link level:
1. At the data link level, db. unified system of physical. addressing
2. The topology should not contain loops; always between the sender and the recipient. the only route.
· On network level (using routers).
A forwarding route is a sequence of routers that connect transit networks.
Routing information in the table may contain:
Information about all existing and available routes
Information only about the nearest routes responsible for further data transmission to the destination node.
Record in table. routing contains fields: network or destination host address, next address. march, auxiliary fields. Methods for filling in tables: manually by the administrator or by means of specials. protocols for collecting route information. On a network, each host contains its own route table.
The choice of this or that route from the table is carried out by the route on the basis of a certain routing algorithm. Algorithms
: static and dynamic (adaptive).
Single- and multi-route algorithms (usually one route is the main, and the rest are backup).
Sibling and hierarchical
Sibling- all routers are equal.
Hierarchical- are used in subnetted networks with their own routing within each layer.
Metrics- indicators used by algorithms to determine the optimal route.
The length of the route, measured in number of hops
Time lag - the time it takes for a packet to travel from source to destination
Communication cost
Reliability indicator (the ratio of the number of errors to the number of transmitted bits)
Bandwidth
Physical distance between nodes
22. Protocols for collecting routing information RIP and OSPF.
The coexistence of various network technologies (coaxial cable, twisted pair (10,100 and 1000 Mbit / s)) sets the task of their joint use in one network. For this purpose, a new type of network devices is used - switches (Switch Ethernet).
Structured LANs are built using workgroup switches, that is, devices with 12-24 10Base-T ports and 1-2 100Base-T ports. These switches provide high-speed access without waiting for each client to access shared resources.
To increase the number of workstations in the network, you can use stacking hubs. Moreover, they can be combined both through common control devices and in a chain. The advantage of the second solution is increased reliability. MAC addresses - addresses of network adapters (Media Access Control)... (10 + 100) - switch designations.
Further development of Switch Ethernet technology led to the emergence of switches that allow connecting workstations to the port, operating at both 10 Mbit / s and 100 Mbit / s. This is achieved using the Auto-Negotiation or Auto-Sensive mechanism. 10/100 switches can be used as workgroup switches or stand alone. Their advantage is the ability to transmit data only to the specified port, without blocking the transmission medium.
Internal address table:
| Address | Port |
| A | 1 |
| B | 2 |
| C | 3 |
| D | 4 |
In addition, each port on the switch has its own memory buffer and table of addresses (MAC addresses) with which it can communicate. This limits the number of WS (collision domain) to which the workstation sends broadcast packets.
Due to the similarity between hubs and switches, 10/100 switches are sometimes referred to as Switched Hubs.
A heterogeneous network is built from subnets operating in different standards, using different technologies. At the same time, they all form a single integrated environment, where a seamless transition from one subnet to another, invisible to the user, is ensured. That is, the heterogeneous network functions as a single system.
Ericsson estimates that by 2018, 30% of the world's population will live in cities and metropolises, which occupy only 1% of the planet's territory. This 1% will generate 60% of the world's mobile traffic, which is expected to grow 10 times compared to 2014. On the other hand, already today about 70% of all data traffic is generated indoors. Comparing these two trends, it becomes obvious that the requirements for network bandwidth in large cities are growing rapidly, as are the expectations of consumers regarding the speed and reliability of data transmission. Telecommunications companies are faced with the challenge of creating networks that would be integrated at various levels, combining different standards and technologies, ensuring a seamless transition from one standard to another, from one technology to another. Such networks should not only combine different standards (from GSM to LTE), but also provide full interoperability between different network layers, as well as networks built on different radio access technologies. It is these networks that are called heterogeneous.
“Since the appearance of base stations of various capacities (macro-micro-pico) and various standards (2G-3G-4G), all networks are in fact heterogeneous,” says Eduard Ilatovsky, a leading expert on planning and development of the Vimpelcom radio network. “Over time, this concept has transformed, and now heterogeneous networks mean a completely different level of integration and interaction of various standards and network layers than it was 10-15 years ago.”
As one of the most illustrative and complex projects of the heterogeneous network, the Megafon company names the construction of infrastructure in preparation for the Olympic Games in Sochi. “On a small territory of the Olympic Park, it was necessary to serve subscribers at large stadiums, the park itself was constantly attended by attendants, guests and participants of the Olympics. All of this was tied to the network in the rest of the city, providing seamless transitions when entering the Olympic Park and leaving it back to the city, ”says Aleksandr Bashmakov, director of infrastructure at Megafon. “Such a fragment of the network gave invaluable experience to the company's engineers, so that similar sections of the network would appear in other cities, primarily in two capitals.”
Heterogeneous networks do more than just allow operators to grow network capacity to meet subscriber needs. Such solutions are also the most economically feasible, since they allow operators to solve local problems without re-investing funds in the development of a macro-network.
Construction of heterogeneous networks
Today, any large city can serve as an example of a heterogeneous network. Ericsson specialists divide the process of creating heterogeneous networks into three stages: improving the macro level, compacting the macro level and introducing the micro level (adding small cells).
The most cost-effective way is to increase the capacity of already built base stations, since sites are one of the main cost items in the construction of a network. In addition, such solutions save time, since there is no need to look for a place to locate new stations. Improvements to the existing network can be realized through the addition of new frequency bands, the use of new radio technologies in the dedicated lower frequency band, the introduction of LTE and the use of various transmit and receive diversity solutions, as well as software performance enhancement of radio access networks.
Ericsson estimates that today HSPA technology still has the potential to increase the capacity and average data rate available to subscribers, while ensuring high reliability of connections and good quality of voice services. Thus, the improvement of the HSPA macro-network, without adding LTE technology, makes it possible to increase its capacity by 4 times (with 4G this figure increases by 10 times).
The next stage in increasing network capacity is the compaction of the macro level. Here, operators' strategies are largely driven by regulatory requirements in a particular market. For example, in North America, the distance between base stations of a macro-network should not be less than 700 meters, while in East Asia and Europe this figure often does not exceed 200 meters. To date, manufacturers offer equipment with reduced requirements for placement density (150-200 meters), which offers to achieve a macro-grid compaction more than 10 times.
After the possibilities of compacting the macro-network have been exhausted, operators are faced with the task of installing micro-base stations in places of the greatest concentration of users and traffic - in shopping centers, stadiums, train stations, and airports. Buildings are particularly challenging, where coverage can also be weak due to high levels of signal penetration through walls, in offices or in remote sites where macro coverage is very weak. In these cases, operators install pico and femto base stations that provide local coverage and actually provide dedicated network capacity for specific users.
Which solution for small cells is suitable in a given situation depends on many factors: the conditions of radio signal propagation, the availability of sites for the deployment of base stations, the availability of transport channels and their quality.
Anna Koroleva, a leading expert on the development of Ericsson mobile broadband solutions in Northern Europe and Central Asia, emphasizes that the introduction of small cells also allows more efficient use of the frequency resource at the operator's disposal: “With proper coordination, there is no need to allocate a frequency resource for small cells, which allows you to serve a large amount of traffic using the same bandwidth and increasing the spectral efficiency of the network as a whole. It also improves the data rate at the edge of the cell, which means the user experience. ”
As a rule, operators install small cells of the HSPA standard, because the greatest load falls on smartphones operating in this particular standard, while the number of devices with LTE support is still small (and is unlikely to increase rapidly in the near future). Another way to expand the network at the micro level is the construction of integrated Wi-Fi networks, which, in addition to improving the quality of communication, also increase the overall performance of the network by transferring part of the mobile traffic over the Wi-Fi network.
In Russia, the concept of small cells has not yet become widespread due to regulatory requirements, as well as technological difficulties associated with the implementation of such projects. However, operators are convinced of the need to develop small base stations of various capacities and different standards to create multi-level integrated networks. “Our portfolio contains developments for applying these solutions both in macro networks when planning a regular network, and for targeted improvement of coverage for corporate clients, and even for entering the B2C market with equipment for creating femto-coverage for small offices and home use,” says Eduard Ilatovsky from VimpelCom. "Which of the developments will be implemented and in what time frame depends, first of all, on the demand for certain services in the market."
Vendor selection
Taking into account the multi-level and multi-standard structure of a heterogeneous network, ensuring the continuous presence of a subscriber in this network, regardless of whether he is connected to it through a macro cell or a small cell, in which standard it works and by what technology, comes to the fore. “As the network becomes more heterogeneous, traffic management, load balancing, mobility between different levels of the network are gaining in importance,” emphasizes Anna Koroleva from Ericsson. "Only a common approach applied to all layers and technologies can achieve network continuity and maximize resource efficiency."
In this regard, the question arises: is it possible to achieve coordination at all levels of the network using equipment from different manufacturers? Logically, we might assume that mono-vendor networks are easier to integrate. Eduard Ilatovskiy from VimpelCom confirms that ideal interaction is possible only in heterogeneous networks built on mono-vendor solutions, however, the use of equipment from a non-main vendor for some network levels is possible. This does not adversely affect the quality of the macro network, while improving the quality of communication inside buildings or in places of local concentration of subscribers.
“For example, in Vimpelcom networks, base stations of different standards can be from different vendors: 2G network from vendor 1, 3G network from vendor 2, and 4G from vendor 3, and in the same networks - pico / femto level can be organized on vendor equipment 4, - says Eduard Ilatovsky. - This solution is quite realistic and workable, however, for correct interaction of all levels and standards of the network, fine tuning of parameters and the availability of an automated network control system based on Self Organized Network solutions, which is also actively used in the Vimpelcom network, are required.
According to him, in the near future VimpelCom plans to switch from a 3.5-vendor model to a two-vendor one. According to Alexander Bashmakov, Megafon also builds networks using equipment from various suppliers, and linking it is a separate technical challenge that the operator's engineers have to face.
Towards 5G
The development of heterogeneous networks not only makes it possible to provide the required capacity and reliability of mobile data networks today. Despite the fact that technological requirements for fifth-generation networks are expected to appear only by 2020, it is already obvious that it will be possible to provide the required highest characteristics in terms of speed, capacity and latency only in a heterogeneous network, one of the fundamental elements of which will be small honeycomb.
“Evolving existing technologies such as LTE and new types of radio access will be part of the future flexible and dynamic 5G system,” says Anna Koroleva from Ericsson. - It will support cross-domain integration and work in several radio access technologies. This system will allow very low latencies to be achieved, and the need to increase capacity will require higher RF bands than are currently used. And so we are convinced that technology integration and multi-layer coordination, which are the core of today's heterogeneous network concept, will become a sustainable platform for further network development and enable operators to maximize the potential and leverage the potential of future technologies. ”
For users, the ubiquitous transition to heterogeneous networks will remain invisible. He will not need to manually switch between standards, access points and different networks. The service provider will do this automatically.