Local UE does not provide an IP BS Manager


The UE does not provide an IP BS Manager. The end-to-end IP QoS bearer service towards the remote terminal is controlled from the GGSN. The scenario assumes that the GGSN supports DiffServ functions, and the backbone IP network is DiffServ enabled. In this scenario, the control of the QoS over the UMTS access network (from the UE to the GGSN) may be performed either from the terminal using the PDP context signaling or from the SGSN by subscription data. 

The IP QoS for the downlink direction is controlled by the remote terminal up to the GGSN. The GGSN will apply receiver control DiffServ edge functions and can reclassify the data (remarking the DiffServ Code Point = DSCP). This may affect the QoS applied to the data over the UMTS access (the TFT may use the DSCP to identify the data to be allocated to the PDP context). 

The end-to-end QoS is provided by a local mechanism in the UE, the PDP context over the UMTS access network, DiffServ through the backbone IP network, and DiffServ in the remote access network in the scenario shown in the figure below. The GGSN provides the interworking between the PDP context and the DiffServ function. However, the interworking may use information about the PDP context which is established, or be controlled from static profiles, or dynamically through other means such as proprietary of HTTP based mechanisms. The UE is expected to be responsible for the control of the PDP context, but this may instead be controlled from the SGSN by subscription. 





3GPP Concept of QoS


3GPP Standard TS 23.207 provides the framework for end-to-end GPRS and UMTS.  The end-to-end QoS architecture is provided in Figure below. It’s describes the interaction between the TE/MT (Terminal Equipment/Mobile Terminal) Local Bearer Service, the GPRS Bearer Service, and the External Bearer Service, and how these together provide Quality of Service for the End-to-End Service. 

It’s also describes IP level mechanisms necessary in providing end-to-end Quality of Service and possible interaction between the IP level and the GPRS level, as well as the application level and the IP level. This covers different architectural aspects of the end-to-end Quality of Service concept and architecture with varying level of detail. In general, other specifications shall be referred to for further details; these other specifications enable the reader to acquire the full understanding of the end-to-end Quality of Service concept and architecture. 



QoS Management Functions in the Network: to provide IP QoS end-to-end, it is necessary to manage the QoS within each domain. An IP BS (Base Station) Manager is used to control the external IP bearer service. Due to the different techniques used within the IP network, this communicates to the UMTS BS manager through the Translation function. The QoS management functions for controlling the external IP bearer services and how they relate to the UMTS bearer service QoS management functions.


QoS Conceptual Model: there are many different end-to-end scenarios that may occur from an UE connected to an UTMS network. The following examples depict how end-to-end QoS will be delivered for a number of scenarios that are considered to be significant. 

The Concept of QoS by ETSI


ETSI standard TS 102 250-2 v2.2.1 (2011) covering the QoS aspects for popular services in GSM and 3G networks. The standard divided into 6 parts book that identified below: 

•  ETSI TS 102 250 Part 1 identifies QoS criteria for popular services in GSM and 3G networks. They are considered to be suitable for the quantitative characterization of the dominant technical QoS aspects as experienced from the customer perspective. 

•  ETSI TS 102 250 Part 2 defines QoS parameters and their computation for popular services in GSM and 3G networks. 

•  ETSI TS 102 250 Part 3 describes typical procedures used for QoS measurements over GSM, along with settings and parameters for such measurements. 

•  ETSI TS 102 250 Part 4 defines the minimum requirements of QoS measurement equipment for GSM and 3G 

•  ETSI TS 102 250 Part 5 specifies test profiles which are required to enable benchmarking of different GSM or 3G networks both within and outside national boundaries. 

•  ETSI TS 102 250 Part 6 describes procedures to be used for statistical calculations in the field of QoS measurement of GSM and 3G networks using probing systems.  

General Consideration: ETSI identifies QoS criteria for popular services in GSM and 3G. They are considered to be suitable for the quantitative characterization of the dominant technical QoS aspects as experienced from the customer perspective. The criteria are described by their name and a short description from the customer point of view.

Phases of Service from the Customer's Point of View 


Figure shows different phases (Quality of Service aspects) during service use from the customer’s point of view. The five QoS aspects are: 

1.Network Availability: is the probability of a telecommunications service that can be offered to customers through a network infrastructure. 

2.Network Accessibility: probability that users can register on the network to be successful so that the network can provide telecommunication services. Network can only be accessed when it is available to the user. 

3.Service Accessibility: probability that the user can access the service you want to use., If the customer wants to use a service, the network operator should provide him as fast as possible access to the service 

4.Service Integrity: describes QoS while using the service and contains elements such as the quality of the content being transmitted, such as sound quality, video quality, and the number of bits transmitted error in the file. Service integrity can only be calculated if the service is accessible to success. 

5.Service Retainability: Service retainability describes the termination of services, in accordance with or against the will of the user. Explains how to end or terminate a service, whether or not the will of the user. Examples of service retain ability parameter are call cut-off ratio or the data cut-off ratio. 


Grade of Service


ITU-T Recommendation E.771 proposes network Grade of Service (GOS) parameters for current and evolving land mobile services. These parameters are defined, and their target values specified, assuming that the network and the network components are operating in their normal mode (i.e. are fully operational). Further, the parameters and their target values assume normal (as opposed to distress or emergency) traffic. 

In this Recommendation, the following traffic GOS parameters are specified for mobile circuit switched services: 

•Post Selection Delay: defined as the time interval from the instant the first bit of the initial SETUP message containing all the selection digits is passed by the calling terminal to the access Signaling system until the last bit of the first message indicating ccall disposition is received by the calling terminal (ALERTING message in case of successful call). 

•Answer signal delay: defined as the time interval from the instant that the called terminal passes the first bit of the CONNECT message to its access Signaling system until the last bit of the CONNECT message is received by the calling terminal. 

•Call release delay: defined as the time interval from the instant the DISCONNECT message is passed by the user terminal which terminated the call to the access Signaling system, until the RELEASE message is received by the same terminal (indicating that the terminals can initiate/receive a new call). 

•Probability of end-to-end blocking: defined as the probability that any call attempt will be unsuccessful due to a lack of network resources. 

•Probability of unsuccessful land cellular handover: defined as the probability that a handover attempt fails because of lack of radio resources in the target cell, or because of a lack of free resources for establishing the new network connection. The failure condition is based either on a specified time interval since the handover request was first issued or on a threshold on signal strength. 



User Perception of QoS vs Operational Performance in Practical


Why are any differences between the results of measurements of QoSE (QoS Experience by the user) and QoSD (QoS Delivered by the provider), whereas the measurement of QoS and network performance are not contradictory? 

In practice, many factors that influence the customer's perception of the QoS service they received from the provider. 

In general, the perception of the customer is to compare the quality of service that they feel with the quality they expect. Customer expectations are influenced by the rates they pay and the information that they know from the media and from books. In general, if a customer feels an expensive, then their expectations for service quality is high as well. 

Provider of telecommunications equipment owned or rented, and operates with the standard of performance they called KPI (Key Performance Indicator). The better prepared KPI, and the more realistic service rates, the correlation between customer expectations for QoS performance telecommunications systeM, will increase. 

To better understand the expectations of its customers, the provider must have good customer service. Customer service should be a very good understanding of operational performance measured through Key Performance Indicators, as well as understand the relationship between customer complaints with performance indicators. 

The task is customer service is two-way. On the one hand, they should be able to answer customer complaints properly, according to the technical conditions of operation. On the other hand, they should be able to give direction to the company, the translation of the customer's wishes into technical performance criteria. 

Providers that are less, in general, ignore the customer service. As a result, customers will be frustrated. Customers have been disappointed, because he felt the complaint was not answered correctly. Provider engineers also depressed, because it was already successfully operating the device in accordance with technical standards, but it is still considered bad by the company who read so many reports of customer disappointment. 


QoSE (QoS Experienced by the User)


QoS experienced by users reflect the subjective point of view of a user in certain circumstances they experienced. Customer satisfaction is one of the driving factors for this type of QoS. In general, QoSE described in nontechnical parameters. Telecom service providers can measure the level of QoSE by conducting a survey to its customers or to seek advice and input from them. At this stage, a user combines personal experience with the expected technical quality of the service it uses. In addition to technological aspects, there are several other factors that affect the level of QoSE. Some of these factors such as starting from the signing of the contract between the user and the service provider, the service provider the ability to handle probleM faced by customers, and the overall relationship between the customer and the service provider. Thus, it can be concluded that QoSE quite difficult to measure because there are several factors "hidden" are not easy to identify. 


QoSD (QoS Delivered by the Service Provider)
QoSD reflect the level of QoS that has been successfully achieved by the telecom service providers. QoSD can test the ability of a telecommunications service provider to deliver the promised QoS.

Radio Access Network - RAN


Here’s where we all look at the radio network. It is the most expensive part of the network. It will have many parts and pieces and could incorporate even more parts as the network matures. Let’s start by explaining that RAN means “Radio Access Network,” and it will have everything outside between the core and the end user. Most people just think of the radios, but the network is more than just radios and core. It is a complex system of connections that need to talk to each other and the core and the user’s equipment, the UE. The UE could be a smartphone, a laptop, a device in a meter or a video camera, or anything that can connect to the network. Don’t limit yourself to thinking it is just LTE because it could be Wi-Fi or another type of wireless format. 4G is a collection of high-speed formats and 5G will only add more formats and complexity to the network. It’s something that you need to be aware of when moving ahead. Although Wi-Fi never panned out as the carriers had hoped, it is still a major part of the network for offload. 

Remember that this book is about deployments. We’re not diving too deep into the architecture. The heart of the RAN is the BTS, base transceiver station. The radio itself. The eNodeB is much more advanced than the radios of old. It could be any spectrum, but to give you an idea of what is in it I made a drawing below that is typical of today’s BTS. 

Remember that there is more to the BTS than just receive or transmitter. It is also a router that connects the backhaul which could have a microwave. The BTS also has batteries to survive outages. Power backup will be in most macro, and small cells Wi-Fi usually won’t have power backup. Now that we have 5G you will also see servers at more sites to support cloud and edge computing. We need the radio heads at macro sites and antennas. Today’s macro BTS have separated the RF from the controller. It is the evolution that has made things so different. Small cells, on the other hand, are an all in one unit. 


TDD and FDD Formats


There are two technologies for LTE. For LTE, they have FDD and TDD which both are viable options. Both are viable options. They are both used by carriers in the USA although FDD has been the choice in the past.  

·       What is FDD? FDD – Frequency Division Duplex is something that was used commonly in 3G. It’s paired spectrum with an uplink band and a downlink band in their specific spectrum. For 1G, 2G, and 3G this was common so you could have a talk and receive channel in the system. There is a guard band in between the transmit band and the receive band. FDD was very popular with GSM and CDMA. It is very difficult to take advantage of MIMO antenna technology in FDD compared to TDD.  

·       What is TDD? TDD – Time Division Duplex is where there is one large piece of spectrum used for uplink or downlink. Any part or percentage can be assigned to be the uplink or downlink. If you have 20MHz of bandwidth available, then you’re not locked into 10MHz up and 10MHz down like FDD. Instead, you have full control over how much goes up and comes down. The downside that some carriers had was the timing of the spectrum, and it's higher bands that have this. However, Wi-Fi spectrum is pretty much all TDD, and it works quite well for data. On the other hand, WiMAX used TDD, and it seemed to be taking off but it never fully blossomed and was cast aside for LTE. TDD makes MIMO technology easier to use because it is all in one band. 

So, what can LTE do? It can do both, and it does do both. Just not the same equipment. You could have equipment do either LTE-TDD and LTE-FDD. Both are released commercially as well as part of the 3GPP standard. When you look at the deployments, it helps to know which format will be deployed. You see, FDD may need two antennas or a combiner to work on a tower. While TDD is all in the same spectrum and the same antenna is used for both transmit and receive. The way that today’s radio heads work it isn’t much of an issue anymore because they can handle the formats quite well. In 2016, you still can’t run them together in the same radio head, although the OEMs are working towards that functionality. Antennas are being designed to run both together by adding more ports and more weight to the antennas. 


Note that Wi-Fi is TDD and ZigBee is TDD. Most Bluetooth is TDD. TDD appears to be the choice moving forward. Most 2G and 3G systems were FDD, and they are being phased out. 

Carriers are learning that when everything becomes truly digital in IP format that it will matter less and less for the BTS, but antennas and spectrum efficiency become more important. As of 2016, most of the carriers already have implemented VoLTE into their main networks, all except maybe Sprint who was still relying on CDMA to carry the voice. The carriers know that when they convert VoLTE, it should be the last step to dismantling the 3G networks, saving them money in the long run by retiring 2G and 3G systems. 


4G spectrum, soon to be part of 5G Spectrum


The spectrum is whatever they could get from the FCC in the USA. They get it from the spectrum auctions that the FCC holds. There is always a need for more although some carriers have yet to deploy all of what they have. With 3G they could use smaller swaths of bandwidth. 4G changed that, and 5G will only make them want more. 

Spectrum is tough to show because there is 4G spectrum for auction here in the USA. I realize that spectrum goes to the highest bidder, (in my opinion small businesses suffer). However, the rush to get spectrum has diminished by the carriers learning to make the most of the existing spectrum. While the bands are small, they have been using something called carrier aggregation to combine spectrum bands to look like one big pipe, which is awesome. The OEMs have worked to put together 2 or more bands so that they look like one big band making the end user happy with more throughput.

In the USA, there are many bands. 

•710 to 716MHz paired with 740 to 746MHz used by AT&T 
•746 to 757MHz paired with 776MHz to 787MHz used by Verizon Wireless 
•806 to 866MHz and 869MHz which belongs to Sprint, this is the old Nextel band. 
•1710 to 1785MHz and 1805 to 1880MHz is T-Mobile AWS spectrum. 
•1850 to 1990 MHz is Sprint FDD spectrum. 
•2.5GHz to 2.7GHz is Sprint TDD spectrum. 
•More and more, it would take some time to break them all out. So much spectrum is out there, and the carriers are grabbing what they can. 


5G Network Slicing


Network slicing is 5G’s way to get you everything. You see, one network will not provide all services for everyone, so they have 5G which will encompass many networks, wireless networks, into one big network. You can’t do everything with one wireless network. Like Steven Wright says, “You can’t have everything. Where would you put it?” If you had one network, it would not be efficient enough to serve all the devices on it. You want a network that works. Otherwise, you have a notwork because it does not work! Most IOT devices don’t need broadband. Most smartphones need mobile coverage. Most laptops need broadband. Most gamers need massive broadband to get the VR to work. Each specific group has a different need. Wouldn’t it be nice if you could have several different wireless networks and have them all go into one core and share resources? Well, 5G came up with network slicing so we can do just that!
The research on network slicing showed me one thing that this is a fancy way to say different networks all connected to a common core. I think this term is interesting, but if you are in IT, then you know that you could have multiple networks, virtual or separated, all sharing the same backbone or even the same physical network. The way I see it, it is all about the RAN! Let’s explore why. 


Well, in 5G, it is not much different. The big difference is that you could have a wireless network dedicated to a specific service. What this means is that when planning a network, in this case, a RAN network, make sure you know what the application will be so that you can plan accordingly.

Think about the different markets 5G will be serving. It could be autonomous cars, virtual reality, or tons of simple IOT devices. Each system will have different need and purpose. The goals are not the same for each. Therefore, they should not all share the same network. So, for the 5G network to include them all, they came up with a cool term like network slicing. The reality is that they will all be different networks that could be sharing the same core or even backhaul. We are creating a way to share resources and build in efficiencies.

We’ll get into why in a few minutes, let’s look at how they will work together first. It’s all about sharing of resources. Think of the HetNet, (Heterogeneous Network) and how we had small cells working with Macrocells and Wi-Fi all working together as one network. Now you have multiple networks all working independently, yet, connecting to the common core.

Which resources are shared in network slicing? The backhaul and the core but also routers and servers and possibly even cloud resources. The key to getting latency down is to rely on the cloud. However, the end user will determine which network will be used and how it will be utilized. The way I see it, from a wireless viewpoint is that the device will need to have a wireless network that fits the needs. In other words, virtual reality with need low latency and very high bandwidth to work properly. Autonomous cars will have very low latency but lower bandwidth needs. IOT devices will have medium latency but very low data rates, and they will not be listening to the network all the time like the other 2, they will only listen to the network on a need to know basis. 

The examples above show us that there will be a need for specific wireless networks to serve each purpose. The common denominator will the core. The core will need to know how to process each part of the network. Making the major carriers happy that they have resource sharing capabilities to save costs. They want to reuse as many resources as possible. Device manufacturers will continue to improve devices and battery life. 



Why Narrow Bandwidth systems in 5G?


Narrow band is for IOT devices. You see, with LTE and Wi-Fi, they tend to be on the air all the time which means the device, a smartphone or your laptop, will be listening and processing data all the time. With IOT devices, they don’t need to talk all the time. They could be pinged once a day or even just talk when they have something to say. 

While there are several reasons, the main one is battery life. If it is talking all the time, then the power draw is constant and high. Broadband kills any battery because it is talking all the time. To get a 10-year battery life, you need to plan when or how it will talk and listen. You don’t want it drawing on that battery 24/7 because it’s listening and processing data. Think about your laptop and how the Wi-Fi will drain the battery life, just like the display. These are the main draws of power. Well, with many IOT devices there is no display, so they only massive power draw is the radio. If you can have the radio go to sleep until it is needed or to wake up at a time of day, then the battery will last a very long time. 

For example, if you have a water sensor or a gas meter or a water meter, three devices that could be mounted where there is no available power source, you need to make sure that battery will last a very long time. Each device will have a different function.  

•The water sensor may only wake up to send a beacon to let the system know that it is alive and working unless there is a high-water alarm, then it will send out alerts. This way the battery will only work when it must. 

•For the metering, gas or water, it doesn’t need to send information all the time. Only maybe once a month or when it’s queried. It may send information of the usage is extremely high to let people know that there is a massive draw on the product measured. This way the battery will last a very long time, and the company deploying these devices will not need to run power to everything. 

These are just a few examples of how the narrowband will be a slice of the 5G network.

Why the need for 5G Low Latency?


The key to true 5G high bandwidth needs as well as low bandwidth needs. The quick response for most devices will be needed so that applications can “talk” as close to real time as possible. You may have seen the RTC, Real Time Communication, a term tossed around RTC is where the device needs to react very quickly, and there is little time for delay. For instance, self-driving cars., They must process the data, so when they communicate with the devices around them, they need to have as little delay as possible. I am talking microseconds, not milliseconds. Why? Because they still need time to process the data. 

Self-driving cars won’t just talk to the network, but they will be talking to cars around them, “looking” all around them, driving the car, making thousands of decisions every second, millions every minute. Deciding how to prepare for the road ahead, the environment around them, and what’s the next move. They will always be concerned about what’s next outside the car and inside the car.

Therefore, the communications system must talk quickly, hence, low latency.

What Applications will 5G have?


For one it will have all that you do now on 4G, internet connection, all the apps, all the things you’re doing now that you feel you can’t live without. 

The new applications will push the network beyond the limitations that we know today and into virtual reality and IoT connections and more streaming video that we could not have before. 

One more thing that is driving it? Vehicle to Vehicle communications. It is the thing that we expect to change the way we live. Vehicles that can communicate with each other to make the chances of accidents lower than ever, in theory anyway. It is also pushing the limits of driverless cars. We are hoping that someday in the next five years that driverless cars are commonplace. Can you imagine that? It would reduce the chance of death on the highways and roads in general. The network will be responsible for all of this, albeit more than 5G but the network reliability, latency, and speed. This falls under the Internet of Things, IOT.

In the enterprise, we will have real-time reporting of KPIs, or stock trades, or horse races that we can get real-time results on, even see the action take place in real time.

Sporting events can offer you the best seat in the house at your home, or at a bar or even at any remote location by showing the game in virtual reality. Think about that, watching the game, be it American Football, Soccer or football, baseball, rugby, cricket, or any Olympic event as if you were in the stadium. The only thing you won’t get is the smell of the arena or someone spilling beer on you. Invite me over, and I can take care of the beer spilling part. 

What about smart cities? Suddenly we pushed cities into the idea that they can see all the city at any time. I know you’re thinking that big brother is watching, but what if big brother was looking at traffic patterns, accidents, traffic delays? They may be able to help or at least report it so that you know to go a different way and in real time. They would also see major potholes in the road and warn is that it is ahead. 

The smart home will go to the next level. When the new networks come online, we should have improved battery life with greater efficiency so that we can take the devices out of the home, away from power, and rely on batteries for months instead of days. We could track pets, bikes, anything that we leave outside for a fraction of the cost it would take to do it now. 

Health services for taking medicine and tracking health conditions will improve, we see it now, it will go to the next lever for real-time reporting anywhere and anytime. 

Drones are always brought up, but the network for drones should be large. With drones, they may have a near field 5G connection that would allow them to control and upload and download data. It is the network that will enable these devices to get their information, but they will need to be autonomous at some point. The network can give them the updates and information they need for the flight path but they need to be able to talk to each other in the air, this is where a small 5G mm-wave network can come in handy. 




Controlling Telecom in a Centralized Way


Telecom management is one area where centralized control tends to generate better results. Of course it isn’t an absolute truth; multinational companies must balance the benefits of centralized management against the difficulties of managing infrastructures in different countries with different languages, currencies, and cultures.  

It is our experience that policies and standards should be defined globally as far as possible. This creates an environment where teamwork and cross-regional support are possible, greatly enhancing the efficiency of the human capital deployed across the organization.  Here we emphasize the need to have unified inventory databases, processes, and technological standards. Sometimes, several arms of a large organization spread around the world do not understand the benefits of unified policies and standards. Usually, the telecommunications team in each country tends to believe that its own ways are the best, but anyone who has managed a multinational telecommunications area knows that having standards is better, even if they are not going to be optimal in every environment. When telecom management is centralized, it leads to the following benefits: 

• better prices (usually due to global negotiation, where the full weight of the organization is brought to the table, yielding better discounts) 
• better control (when only one group is responsible for telecom resources, it usually reduces problems such as overcharges, overlaps, and having unidentified resources or resources that are not used) 
• lower operational costs (when headcounts are reduced, there is a consequent reduction in personnel costs) 

Centralizing control usually enables the organization to identify its telecom expenses. That fact alone is usually enough to justify centralization, because it shows how much telecom represents within the IT/infrastructure budget and keeps the subject on management’s radar.  
In more general terms, we have to keep in mind that telecom is a logistic system, and as such, the whole may be more than the sum of the parts. 

It would be interesting to insert a caveat into the argument here that centralized management doesn’t necessarily mean a centralized operation. If you have the right tools, you may be able to control and contract in a centralized way and yet keep the operation distributed, enabling different telecom teams to operate in different countries, for example.  

This is feasible, as long as you manage to make all teams use the same management tools, under a defined hierarchical framework. That means that the local telecom teams may have some autonomy to contract telecom resources (the ones not covered for the worldwide contract, for example), but they have to include each contract and resource in a corporate telecom management tool in such a way that headquarters can see all the telecom expenditures and all resources contracted in all countries. The local teams will see only their own expenditures and resources.  

Therefore, we may divide the term “centralization” into two types: financial and technical. Even if operational aspects force technical decentralization, financial centralization remains crucial. The centralized telecom management has to keep track of what is contracted and how much it is costing.  Financial centralization refers to the following:

• centralized resource inventory (including data, voice, and mobile resources) 
• centralized contract inventory (including voice, data, mobile services, and maintenance) 
• centralized telecom bills (even if received in different countries, all bills would be included in a common tool in a standardized framework, allowing centralized control) 
• centralized billing system 
• centralized bill auditing process (at least in a country basis) Technical centralization refers to the following: 
• centralized help desk for telecom issues 
• centralized point of contact with the telecom providers 
• centralized point of contact for equipment maintenance 
            • centralized network operational center (NOC) 


Telecom Made Simple

Related Posts with Thumbnails