What is 5G? I am currently in the process of writing a book and will be working on some ideas in my blog. My book will be called “Viva5g” and in my book I will have several series one for for “Entrepreneurs and Executives” and another series of books written into more technical depth for engineers and experts.
Where is 5G derived from? Answer: Standard Bodies
5G is what is called the fifth generationwireless communication technology or standard. The current wireless communication standard is called 4G or the fourth generation. Therefore, there was a third and a second generation, or 2G and 3G? Who wouldn’t recall purchasing an iPhone 3S with 3G enabled?All of these wireless communications generations are nothing but ways for carriers to marry or meet and match certain “terminology” to a set of standard or standards, and consumers to identify what they are getting into. Not only consumers, but investors, media, and even their own workforce. It is a bit complicated to make precise mappings, but I will try to explain it in the following paragraphs.First, you may have seen “5GE” at the top right corner of your AT&T iPhone? If not, this picture depicts mine.Why is it 5G now and my iPhone is a X or iPhone 11?What happens is that behind the scenes, there is an organization called the “3GPP” or the “Third Generation Partnership Project,” that establishes and deals with very complex standards. The 3GPP groups is composed of other standard bodies and all boils down to a group of companies that are 3GPP “members.” This club or “members” get together in multiple committees meet with the purpose to define, literally, to define how things will work or operate into the future.Some people may say on youtube videos that 5G is the devil and may even hint that the 3GPP is part of the “new world order” and is here to control us all, the answer is no!!! That’s just fiction and conspiracy theories from youtubeers. 3GPP is just a group of companies and several other standard development organizations worldwide that have built or developed ways to improve and make things better, faster, higher-performance, and meet to agree on how to get this process done.Let’s start with 4G or the fourth generation wireless network, which we already use, or you maybe using to read in this posting.4G corresponds to a mapping made to the 3GPP organization via a set of “Technical Specifications” has labeled as Releases 8, 9, 10, and 11, and maybe 12, and 13, whereas 5G corresponds to a set of “Technical Specifications” improved and changed that are labeled Release 15 and 16.For example, 3GPP Release 14 includes many new concepts not found din Release 8, those are: Internet Of Things, Vehicle-to-Everything, Radio Improvements, etc, as shown in the following screen shot.In between those release, let’s sy Release 15 and Release 11, we find a gray area where 4G ends and 5G begins. In fact, many features from 5G are found in releases that are supposedly part of 4G, as well as new features for 5G appear in newer releases only,. That’s is why some companies like AT&T have called this process “5GE,” and presented us with a “logo” or an icon, that is shown int your iPhone’s screen.This simple icon caused Verizon to file a lawsuit to AT&?T for doing so and t is not clear what 3GPP releases AT&T refers to 5GE..Going in to more detail, the 3GPP organization defines itself in the 3gpp.org website as follows :
The 3rd Generation Partnership Project (3GPP) unites [Seven] telecommunications standard development organizations (ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC), known as “Organizational Partners” and provides their members with a stable environment to produce the Reports and Specifications that define 3GPP technologies.
In fact, the communications standards created by the 3GPP cover multiple technologies for instance: Radio Access Networks (RAN), Services & Systems Aspects (SA), and Core Networks & Terminals (CT), as well as many other aspects.3GPP is hence, as you can imagine, a complex body and I won’t go into more details. In my opinion, one of the main things that are distinguished by 4G and 5G is are the core components, handover, use of frequencies, and physical layers, more importantly the “softwarezation” and the use of software-based technologies at a much higher frequency band, which leads to higher bit rates.In essence, in today’s 4G and 5G core networks a concept has been introduced and it is called, “Network Function Virtualization” or NFV, which brings the cloud to telecommunications systems together, making all the changes in software, and much less in hardware. Obviously, there are servers, CPUs, GPUs, and all that working to manage all signals, but that is multi-purpose and can be easily upgradable, as in the past, it was not.
Cloud Computing and Open Networks
As you may expect, NFV brings as a main feature, the virtualization of all network components done in multiple instances or containers, which is popularly called, the cloud.
It’s easy to confuse virtualization and cloud, particularly because they both revolve around creating useful environments from abstract resources.RedHat.com site
I agree with RedHat’s comments, however, virtualization is what has made all cloud computing concepts possible, and it is clear that has come to optimize and improve 4G and hence 5G systems.Therefore as the cloud makes its way to 4G and more to 5G, we will find new terms that were not part of telco’s, one of those is “Open Network.” Fir instance, “Open Mobile Evolved Core,” or an open “Core” network, which is not necessarily “open source” but what that means is that APIs are used just like in any cloud-based environment, and things or components can be easily interconnected.Therefore, under this model, all 4G and 5G network components and hence the entire 5G network runs in the cloud with servers executing multiple instances of machines or containers.For example, there are any standardized 4G components that include Mobility Management Entity (MME), Service Gateway Control (SGW-C), Packet Gateway Control (PGW-C), Policy Charging Rules Function (PCRF), among some others, those now run as server instances.For instance, the MME or Mobility Management Entity, is a server or cloud of servers entities that handles mobility and tracks the mobile terminal in the network, assisting the UE or User Equipment, or your mobile phone with handover and selecting the right cell to move to, as it travels around a different geographical area.In this picture taken from the “Open Mobile Evolved Core,” we cab read that all of these components may all reside in one server, and the network created is virtual, or a software-defined network.
What is then Network Function Virtualization?
In essence, Network Function Virtualization (NFV) is nothing but a way to put all these network elements or components in servers or virtual machines. These virtual machines run in standard VMWare servers or Docker Instances, and now you can introduce standard cloud computing concepts and tools like Open Stack, or Kubernettes, for what is called “Orchestration” or the process of creating and making instances.
Orchestration is the automated configuration, coordination, and management of computer systems and software. A number of tools exist for automation of server configuration and management, including Ansible, Puppet, Salt, Terraform, and AWS CloudFormation. Wikipedia
These virtual machines are the main fabric of the “cloud,” the “cloud” is a set of machines or virtual machines that reside, ultimately still in datacenter and servers, somewhere, but are sufficiently operational from a “File” or an “Image” that can be copied into multiple data centers and operate without issues. These images or files are stored as “containers” or “virtual machines” that are executed in a real or hardware machine, that performs certain task.
Another concept that has existed among “computer science” professionals for many decades, is: moving computation to the “edge” of the network. It seems quite obvious, but it is not.Edge computing is a natural evolution to optimize processes.Historically, in the past, mainframe computers controlled everything with a dumb terminal that displayed what the mainframe processed. Later on, the PC or a Personal Computer evolved and all control was passed down to the PC and some servers, the use of mainframes becomes less important. As progress created the internet, all moved to web and cloud. Now, the control has been passed on to the cloud or a distributed computing system governs what we do and how we do it. Hence, the closer computation is being moved, in proximity your resources are to you, the better, lower latency, falser responses, but also causes problems for the overall system to maintain authentication, caching, and other dependencies.For example, we use “Edge” computing every day, in “content delivery networks” (CDN) that are used by NETFLIX or HULU to stream movies to thousand or millions of homes. The CDN’s r main goal is to move content, music or video files, closer to your local internet link by making multiple copies available closer to the consumer or to the “edge” of a network.This concept is practical when all components use the same protocol, in this case IP or the Internet Protocol. The cloud, and an “ALL-IP” network architecture is found in 4G already, and subsequently is found in 5G.Therefore edge computing is now a more formalized concept, feasible and practical.The major evolution from 3G to 4G is that “all” components in 4G including the core infrastructure in 4G (and obviously in 5G) runs over internet protocols or IP. This evolution is a a major distinction between UMTS, CDMA, and older systems with 4th generation wireless networks.In 5G, IP is also the main fabric for communications, and all signaling data, voice, videos, are no longer using proprietary signaling but “IP.” In the past, UMTS or GPRS, IP was an afterthought and was emulated on top of those proprietary protocols which made the networks slow and expensive to maintain. Obviously, challenges surfaced as 3G moved to 4G including changes in handover protocols, billing, and access to the network.
These proprietary protocols in 2G and 3G were designed to operate “Circuit Switched Networks” and 4G is an all “IP” network designed to operate in a 100% Packet-based Network.
Edge Computing works well with all IP networks and now the way to go. As computation is moves to the “Edge” and as the “Edge” of the network is now a server or a cloud-based component of a a bigger cloud, many new ideas are being formulated and network and computation blend in as what is called today a “Software Defined Network.” In other words, the network is defined virtually by software and multiple networks can be created using the same physical interconnections.Edge Computing is :
Edge computing is a distributed computing paradigm which brings computation and data storage closer to the location where it is needed, to improve response times and save bandwidth.
Edge Computing is now a major area of innovation, for example “Cloud to Cable,” my own patented technology is an “Edge Computing” entity that facilitates caching, which is data storage, and computation which is covered by my own patents, closed to the cable operator. However, the same is true for a 4G and 5G system. I am personally working on how to achieve that at EGLA Research Labs. As a consequence of the use of the cloud, organizations like the “Open Network Foundation” or ONF and others, are looking for ways to standardize how this is done.
The Open Networking Foundation (ONF) is a nonprofit trade organization, funded by companies such as Deutsche Telekom, Facebook, Google, Microsoft, Verizon, and Yahoo! aimed at promoting networking through software-defined networking (SDN) and standardizing the OpenFlow protocol and related technologies. The standards-setting and SDN-promotion group was formed out of recognition that cloud computing will blur the distinctions between computers and networks. The initiative was meant to speed innovation through simple software changes in telecommunications networks, wireless networks, data centers and other networking areas.
Obviously, since now the cloud is powering 4G and 5G, the same standardization and SDN with protocols like “OpenFlow” are now plausible to be used in the network infrastructure of 5G systems. In fact, the fabric of a Software Defined Network, has always been of my own use at EGLA since 2014, when we moved into an “Equinix” data center with the first version of the Mediamplify platform.
Network Slicing and Beamforming
Another concept that has been introduced mostly with 5G is beam forming and network slicing.Network slicing is used to assign an IP Address or a network for your own company or a user, this enhances quality of service to what a user has paid or a company is paid. A separation of virtual networks within the core network is being done for purposes of individualized routing and treatment of user’s traffic.
Network slicing is the separation of multiple virtual networks that operate on the same physical hardware for different applications, services or purposes.
Similarly, as part of the network is “Sliced,” the RF or Radio Frequency or wireless signals are now set to multiple beams.In a way, 5G operates at a high-bit rate, up to Gigabits per second, but at much higher frequency bands, sub-6 GHz or over 6 GHz. At these frequency bands, propagation of signals and physics brings the size covered by a base station to a smaller footprint. In other words, the power levels and noise are not appropriate to establish a link at 1km but are great at 100m, for example, and as opposed to 4G, a sector instead of covering a wide area, covers a few meters of wide. According to Metaswitch site:
“Due to the high propagation loss of the millimeter wavelengths (mmWaves) employed in 5G new radio (5G NR) systems, plus the high bandwidth demands of users, beamforming techniques and massive Multiple Input and Multiple Output (MIMO) are critical for increasing spectral efficiencies and providing cost-effective, reliable coverage.”
Hence, signals are sent from multiple antennas (MIMO) and received by multiple antennas at the phone. This is already being achieved in 4G, at a smaller level, with a a technology called “Carrier Aggregation.”As expected, now that all operations and network is based on software and running on server and virtual machines, just like google cloud, Amazon’s cloud, Azure, Digital Ocean, and EGLA CORP’s cloud-based servers, what can stop “Artificial Intelligence” from being used? The answer is nothing, Artificial Intelligence or AI has been incorporated to work with the network.AI, Machine Learning, and other methods are used for network optimization and use radio resources better, to optimize power management and decrease electricity bills, frequency reuse at the radio-level, handover optimization, and network management with predictive failure detection.The machine learning mechanisms in existence today, can learn from large amount of data logs collected by the telcos’ and are perfectly suitable for cellular networks. The cellular network adapt and generate more data for thousands or hundreds of thousands of base stations that are deployed with millions of users in phones, IOT (Internet of Things) devices, and connected vehicles. The possibilities are endless.Here some AI examples in telcos of what I posted in my show, TECHED.TV.
Low Latency and High Bandwidth
Low Latency is now another factor of great importance for 5G. Low-latency makes robotics and self-driving cars possible. Before, latency was 800ms or a few seconds, let’s in GPRS and CDMA 2000, even UMTS provided latencies of 200-300ms.
Network speeds in 5G will be in the Gigabits per Second
High Bandwidth as expected would be in the order of Gbps or Gigabit per Second.LTE decreased latency to a few tens of milliseconds, but remember that you have to connect to the internet and account for all signaling, which provided an overall latency of 40-60ms, still unsuitable for remote robotics.Since the access or network is sliced, a portion of that network could be allocated to be higher priority and hence decrease latency of the overall access to a few milliseconds, which is now perfectly suitable for robotics and automation, or even self-driving cars, AR, and 5G Gaming.
An X2 Handover is used by LTE as well as S1 handovers for User Equipment (“UE”) or mobile phone mobility.
First and foremost, LTE and 5G both share an ALL-IP Architecture. In the past, in UMTS systems or even older systems, mobility consisted in several layers of IuB interfaces from the RNC to the NodeBs. It is not then an ALL-IP network as expected.
We will not go over Tracking Area Updates and Tracking Area Indentifiers as that’s will be discussed in a later post.
In UMTS, the Radio Network Controller (RNC) manages hundreds of NodeBs, all interconnected using the IuB interface and RNC to RNC uses the IuR Interface. The SGSN or Serving GPRS Support Node is in charge of setting up the internet link or GTP-U tunnel from the SGSN tot he RNC. However the IP link will not move although an you can switch from NodeBs, meaning that any soft-handover or hand-handovers that occur between NodeBs does not affect the GTP-U link form the internet to the UE.
In LTE, on the other hand there are many changes, one is that eNodeB is the end-point of the IP link, making it an ALL-IP network. This change requires some redesign to the protocols and specially the insertion of “hard handovers” plus X2 and S1AP Handover protocols that are now based on IP Mobility. As shown in the figure below, X2 interfaces communicate eNodeBs, and S1 links from eNodeBs to S-GWs and S1-MME to the MME. The MME is the Mobility Management Entity which tracks the UE and does paging, as well as updates.
As you can observe, the X2 Interface is key for handoff as this interface is used not only to detect adjacent eNodeBs but also to exchange information as interference and others.
The X2 handover will be illustrated as follows:
The UE has to be in RRC_CONNECTED stated not in RRC_IDLE where a process called “Cell Reselection” i used. Lets Start with 1-6 Steps:
Step 1: Handover Command
As opposed to others, we will start with the Handover Command, which is the an RRC Connection Reconfiguration Message that contains a field called “MobilityConttrolInfo” this field contains the Physical Cell ID to handover to, as well as a list of neighbor cells with their associated “Cell Individual Offsets” that are used for A3, A5 and other events.
The Handover Command forces the switch from a previous cell to the next cell. Handover can only occur
In this example a Handover Command was issued to connect to eNodeB with ID=1.
This structure can be found in the specification as follows:
Step 2 : Measurement Reports and Events
As we know, LTE specification defines several reports. In this example we will focus on A5 and A3 events, that are programmed using the Cell Individual Offsets and Frequency offsets found in the specification. Since the device is in “Connected State” Layer 3 filtering is applied to the measurements made by the UE.
The filtering algorithm may use different coefficients that the eNodeB sets as default.
Events and measurements
The state machine inside the UE, is configured by the rrcConnectionReconfiguration message to track all the eNodeBs provided and its frequencies, including applying SIB4 blocks that are submitted by the eNodeB to the target.
In a way the eNodeB is predicting the next state to follow reporting A5 and A3 Events to the eNodeB.
STEP 3: HANDOVER REQUEST
Now it is known by the eNodeB that Handover might be required and decides based on all the events or measurement reports, where to Handoff the UE too, and creates a HANDOVER REQUEST to a a Target eNodeB or the one with ID=2.
STEP 4: Allocation of resources in Target eNodeB
Now that Handover Request is moving forward, a setup for tunnels is created to the Target such that all internet traffic will start flowing tot he Target eNodeB with Cell ID =2
Once this is successful a Handover Acknolwedgement is made to the source and packets start going to the Target, similarly, Downlink and Uplink tunnels start being moved to the target eNodeB with Cell ID = 2
STEP 5. Handover Command to Switch to Target
Exactly as in STEP 1, a Handover Command (in an rrcConnection Reconfiguration message) containing the Target ID = 2 is issued from the Source, also at some point another rrcConnection Reconfiguration message will contain all neighbor nodes with its respective Cell Individual Offests and this process will continue.
STEP 6. All is switch to the Target and MME is updated of the move.
As a final step of handover, the path switch is complete which is a process called “Late Path Switch” that generates all traffic to move from the Source to the Target eNodeB in its totality.
You can hire us for consulting for your 4G or 5G Project.
We have R&D Labs with wireless capabilities for IOT, Mobile, Emulation
Hands-on Experts in Computer Communications, Networking, Cloud, Network Function Virtualization (NFV), Edge Computing, Smartphones, and all your strategic intellectual property needs.
BigData Presentation – Radio Frequency / Mobile CDS Intelligent Positioning of RF Cells for 4G/5G
I was invited to FAU (Florida Atlantic University) to present at one of the MBA classes on “Big Data Analytics” and we went over the important concept and examples of MapReduce, Hadoop, Pandas, and sample on how Radio Frequency can be simulated and how Big Data is the key component to process, aggregate, and create dashboards of RF simulations over 3D KML maps loaded from Google Earth/Google Maps. This presentation also covered aspects on how the data is split and can be splitter in multiple GPUs using OpenCL as a framework.