Understanding 5G and LTE - SGi-LAN

SGi-LAN (SGi Local Area Network) is network infrastructure connected to a 3GPP (Third Generation Partnership Project) LTE (Long-Term Evolution) network over the SGi or Gi reference point (i.e., interface) that provides different value-added IP-based services to user data as it flows through the network [3GPP]. These value-added services may include, among others [3GPP, Cisco, Intel-1]:

• NAT (Network Address Translation)
• Anti-malware
• Parental control
• DDoS (Distributed Denial of Service) protection
• Firewall
• Policy and charging enforcement
• Traffic detection (PCEF/TDF; Policy and Charging Enforcement Function / Traffic Detection Function)
• Content Delivery Network (CDN) caches
• Video transparent caching and optimization
• TCP optimization (e.g., to prepare the traffic for the Radio Access Network)
• Shaping traffic with Deep Packet Inspection (DPI)
• HTTP header enrichment to support partner services
• Providing analytics information
• Differentiated charging

In the future, SGi-LAN will also need to address the growing use of encrypted traffic. It might also include different forms of proxies and interworking functions for IoT [Cisco].

In an LTE network, SGi-LAN is positioned between the Evolved Packet Core (EPC; the LTE core network) and a PDN (Packet Data Network) Gateway such as an Internet or IMS (IP Multimedia Subsystem) gateway. A typical SGi-LAN supports several million subscribers [Intel-1].

SGi-LAN is the home of Service Function Chains (SFC), which are not standardized by 3GPP [IETF]. Traditional mobile deployments of SGi-LAN services simply chain all the functions together in a serial and static manner [HPE]. The limitations of this approach include that (1) each function has to process all the traffic, (2) introducing new functions or upgrading existing ones takes a lot of time and effort, (3) a failure of a single function may interrupt the overall service for all subscribers, and (4) the introduction of new services is slowed down.

Service Function Chaining (SFC) makes it possible to dynamically configure user plane traffic to be routed through a chain of network components which provide value-added services [SS]. As an example, traffic of a certain customer may be passed through a protocol optimization component (e.g., for video) or a security function such as parental control. SFC, when combined with SDN (Software-Defined Networking) and NFV (Network Functions Virtualization), can be used to remove the limitations mentioned in the previous paragraph and enable optimal use of data center resources, scalability to cope with ever-increasing traffic, and optimal steering of traffic through SGi-LAN network functions [Intel-2]. In fact, virtualization of the SGI-LAN is often a first step for mobile operators towards network virtualization [HPE]. It allows the virtualized SGi-LAN functions to be run as VNFs (Virtualized Network Functions) on standard x86 server hardware in the operator’s cloud infrastructure.

3GPP is defining the integration of 3GPP policy standards with external policy standards, including policies for SDN controllers. A new interface (St) has been defined between the PCRF (Policy and Charging Rules Function) and a new Service Chain Traffic Controller Function (SCTCF) [Intel-2]. The St interface allows the PCRF to interface to the SFC controller functions in order to provide traffic description filters that enable more coordinated and comprehensive implementation of service chains in SGi-LAN.

But what is the role of the SGi-LAN in 5G networks? According to some visions, in 5G networks, the integration of SGi-LAN network functions will be seamless, and the traditional separation between the network functions placed in the SGi-LAN and other parts of the network will be removed [NOK], allowing the SGi-LAN functions to be either integrated with the basic connectivity and policy enforcement functions, or still be deployed independently in the SGi-LAN. Among other things, this will allow service function chaining to take advantage of edge clouds - virtualized SGi-LAN functions can be placed where needed [AF]. As an example, SGi-LAN functions could be placed on router-based compute blades, on COTS (Common Off-The-Shelf) x86 servers on an edge node (e.g, one could place a CDN cache on an edge node to move traffic closer to the edge), or in a central data center. The placement decision is made by a central orchestration system, which can steer traffic into service chains for example on a per application, subscriber, bearer or device basis, or based on some combination of them.

Finally, SGi-LAN may also be referred to as (S)Gi-LAN, Gi-LAN, GiLAN, vGiLAN (virtualized GiLAN), SGiLAN, Gi LAN or SGi LAN.

References

[3GPP] 3GPP TS 23.203 V14.3.0 (2017-03), Technical Specification, 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects;
Policy and charging control architecture (Release 14)
https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=810

[AF] Designing 5G-Ready Mobile Core Networks, https://www.affirmednetworks.com/wp-content/themes/renden/pdf/5G_Whitepaper_Heavy_Reading.pdf

[Cisco] The Cisco 5G Strategy Series: Packet Core, Transport, and Identity Management, http://www.cisco.com/c/dam/en/us/solutions/collateral/service-provider/ultra-services-platform/5g-strategy-series.pdf

[HPE] Virtualization of Gi-LAN Functions, https://www.hpe.com/h20195/v2/GetPDF.aspx/4AA6-3417ENN.pdf

[IETF] Service Function Chaining Use Cases in Mobile Networks, draft-ietf-sfc-use-case-mobility-07, https://tools.ietf.org/html/draft-ietf-sfc-use-case-mobility-07

[Intel-1] Etisalat and Intel – Virtualizing the Internet Gateway Gi-LAN for Flexibility, https://builders.intel.com/docs/networkbuilders/Etisalat-and-Intel-virtualizing-the-internet-gateway-Gi-LAN-for-service-flexibility.pdf

[Intel-2] Gi-LAN and Dynamic Service Function Chaining for Communications Service Providers, https://builders.intel.com/docs/networkbuilders/Gi-LAN-and-dynamic-service-function-chaining-for-communication-service-providers-ra.pdf

[NOK] 5G - a System of Systems for a programmable multi-service  architecture, http://resources.alcatel-lucent.com/asset/200012

[SS] The Road from EPC to 5G, https://www.slideshare.net/AlbertoDiez4/mobile-plots-from-epc-to-5g a

What Is 5G Xhaul?

To put xhaul (also known as crosshaul or X-haul) into a context, let's start from the domains of a mobile network. These include [5GPPP-1]:

• Radio Access Network (RAN) domain
• Transport network domain
• Core Network (CN) domain
• Other domains, like the OSS/BSS domain for supporting telecommunication services and application domain for providing services [NM]

Xhaul is a concept related to the second domain on the list above, that is, the 5G transport network. The transport network consists of the backhaul of radio base stations and fronthaul of remote radio units (a.k.a., Remote Radio Heads, RRHs). In the case of LTE, fronthaul refers to the high-bandwidth (optical [EX]) transport links that connect the RRHs to a BaseBand Unit (BBU) pool in the LTE C-RAN (Centralized Radio Access Network) architecture. Backhaul is the IP (IP/MPLS [EX]) network from the centralized BBUs to the LTE Evolved Packet Core (EPC). Finally, xhaul could be defined as a common flexible transport solution for future 5G networks, which integrates the fronthaul and backhaul networks with all their wired and wireless technologies into a common packet-based network that is under SDN (Software Defined Networking) based and NFV (Network Functions Virtualization) enabled common control [ER].

This SDN-controlled integrated backhaul and fronthaul transport network (i.e., xhaul) aims to enable a flexible and software-defined reconfiguration of all network functions in a multi-tenant and service-oriented manner [5GPPP-2]. The xhaul transport network is envisioned to consist of high-capacity software-defined switches and heterogeneous transmission links (e.g., fiber or wireless optics, high-capacity copper, and wireless millimeter wave links). These links will interconnect RRHs, macro and small cells, and distributed physical and virtual 5G RAN and core network functions (i.e., Physical Network Functions (PNFs) and Virtual Network Functions (VNFs)) hosted on in-network cloud nodes.

The vision is that Xhaul will turn the fronthaul and backhaul into a unified IP network [Cisco]. The unified approach will provide support for key timing issues like phase and frequency, and increases in network scale. The xhaul transport network needs to be application-aware and automated. Xhaul will be a key enabler for network slicing. Slicing will construct xhaul network resources based on SLAs (Service Level Agreements) and the needs of the application. The xhaul resources will be automated, model-driven and programmable (i.e., SDN-based) to be able to deal with real-time service transition and specific latency demands.

References

[5GPPP-1] 5G PPP Architecture Working Group, View on 5G Architecture, https://5g-ppp.eu/wp-content/uploads/2014/02/5G-PPP-5G-Architecture-WP-For-public-consultation.pdf

[5GPPP-2] 5G Crosshaul, the integrated fronthaul/backhaul, https://5g-ppp.eu/xhaul/

[Cisco] Are you ready for 5G xHaul? http://blogs.cisco.com/sp/are-you-ready-for-5g-xhaul

[ER] X-haul, fronthaul and backhaul network research, https://www.ericsson.com/research-blog/5g/x-haul-fronthaul-and-backhaul-network-research/

[EX] Overcoming Network Failures in C-RAN for 5G-Readiness, http://www.exfo.com/corporate/blog/2017/overcoming-network-failures-in-c-ran-for-5g-readiness

[NM] LTE Network Architecture: Basic, http://www.netmanias.com/en/post/techdocs/5904/lte-network-architecture/lte-network-architecture-basic

What is 5G NR (5G New Radio)?

5G NR (5G New Radio) is a wireless network standard for the next (i.e., fifth) generation of mobile networks. 3GPP (Third Generation Partnership Project) is expecting to complete the NR specification in two phases. Phase I will be completed in June 2018 (3GPP Release 14) and Phase II in December 2019 (3GPP Release 15) [3GPP-1].

NR aims to provide a single technical framework addressing a broad range of use cases, requirements and deployment models, including [3GPP-1]:

• enhanced Mobile BroadBand (eMBB) – MBB is the use case for which the previous generations of mobile networks have primarily been optimized
• Massive Machine Type Communications (M-MTC), which is also known as Massive Internet of Things (M-IoT)
• Ultra-Reliable and Low-Latency Communications (URLLC), which is also referred to as Critical MTC (C-MTC) or Critical IoT (C-IoT)

These use cases place very different requirements on NR and the whole 5G system. Critical IoT (e.g., remote controlled machinery, cloud robotics, and self-driving cars) requires extremely low latency and extreme reliability, whereas massive IoT (e.g., large network of stationary weather sensors) has less stringent requirements on latency, reliability, bandwidth and mobility, but requires extreme coverage (enabled by low frequency spectrum). eMBB (e.g., Ultra-High Definition video streaming) requires extreme bandwidth (enabled by millimeter wave spectrum and small cells) and mobility but does typically not require as low latency or reliability as critical IoT.

NR Capabilities

NR aims to meet the 5G requirements outlined by ITU-R (International Telecommunication Union Radiocommunication Sector) for IMT-2020 (International Mobile Telecommunication system) in the IMT-2020 recommendation document (see reference [ITU-R-1]). 3GPP will submit NR as a candidate technology to the IMT-2020 process. Initial submission is expected to happen by June 2019, and a detailed specification will be submitted by October 2020. The IMT-2020 capabilities include [ITU-R-1]:

• Peak data rate of 20 Gbit/s per user or device. This is the maximum achievable data rate under ideal conditions. For IMT-Advanced (i.e., LTE), the peak data rate is 1 Gbit/s.
• User experienced data rate of 100 Mbit/s in wide-area coverage cases (e.g., urban and sub-urban areas). This is the data rate that is available ubiquitously across the coverage area to a user or device. In hotspot areas (e.g., indoors), the user experienced data rate is expected to be 1 Gbit/s. For IMT-Advanced, the user experienced data rate is 10 Mbit/s.
• Latency of 1ms. This is the contribution of the radio network to the end-to-end latency (i.e., the over-the-air latency). For IMT-Advanced, the latency is 10ms.
• Mobility for up to 500 km/h (e.g., when the user is on a high-speed train). This is the maximum speed at which a defined QoS and seamless transfer between radio nodes can be achieved. For IMT-Advanced, the maximum speed is 350 km/h
• Connection density of one million devices per square kilometer. For IMT-Advanced, the figure is 100,000 per km^2
• Network energy efficiency of 100 times higher than for IMT-Advanced
• Spectrum efficiency that is three times higher than for IMT-Advanced
• Area traffic capacity of 10 Mbit/s/m^2. This is the traffic throughput served per geographic area. This represents a 100x improvement over IMT-Advanced
• Spectrum and bandwidth flexibility, such as the ability to operate at different frequency ranges, including higher frequency (e.g., millimeter wave) and wider channel bandwidths than today
• Reliability, that is, the capability to provide a given service with very high availability
• Resilience, which refers to the ability of the network to continue operating for instance after natural disturbances
• Security and privacy, including encryption and integrity protection of user data and signaling, end user privacy (e.g., preventing unauthorized user tracking), and protection of network against hacking, fraud, Denial of Service (DoS), man-in-the-middle attacks, etc.
• Long operational lifetime, which is important for IoT devices that require a very long battery life (e.g., more than 10 years)

NR will achieve the capabilities listed above through [NOK-1]:

• Massive densification of small cells – due to constraints on spectrum efficiency (Shannon’s law) and use of higher frequency spectrum, cell densification is needed for 5G to be able to deliver the required data rate and capacity improvements [NGMN]
• Additional spectrum (see below)
• Increased spectral efficiency, for which a key technology will be massive MIMO (Multiple Input Multiple Output)

Spectrum

NR will be using both more traditional cellular access bands below 6 GHz and large amounts of spectrum available above 6 GHz [NOK-1]. In fact, NR should be able to use any spectrum band ranging at least up to 100 GHz that may be made available for wireless communications [3GPP-1]. NR will be using [QC-1]:

• Low bands below 1 GHz to achieve longer range for example for mobile broadband and massive IoT. The could include for example 600 MHz, 700 MHz, and 850/900 MHz
• Mid bands from 1 GHz to 6 GHz to provide wider bandwidth for example for eMBB and critical IoT. Examples include 3.4 - 3.8 GHz, which includes also the CBRS (Citizens Broadband Radio Service) bands, 3.8 - 4.2 GHz, and 4.4 – 4.9 GHz.
• High bands above 24 GHz, which will be helpful for reaching extreme bandwidths. Examples include 24.25 – 27.5 GHz, 27.5 – 29.5 GHz, 37-40 GHz, and 64 – 71 GHz.

NR will be leveraging a combination of licensed, shared, and unlicensed spectrum with a goal of having a unified design across all of these spectrum types and bands, including both below and above 6 GHz.

The frequency bands between 3 GHz and 30 GHz are known as centimeter wave (cmWave) bands, whereas bands between 30 GHz and 300 GHz are known as millimeter wave (mmWave) bands. At mmWave frequencies, the radio propagation and Radio Frequency (RF) engineering is different from the traditional sub-6 GHz cellular access bands. For example, diffraction (short radio waves are affected more by obstacles than long radio waves [LR-1]), foliage (radio signal running into leaves on a tree), and structure penetration losses (penetration loss of radio waves for example through walls of a building tends to increase with frequency [ITU-R-2]) are higher at mmWave frequencies [NOK-1]. However, mmWave frequencies are still similar to sub-6 GHz frequencies when it comes to reflections and path loss exponents. mmWave’s reach and ability to support mobility can be extended from Line-of-Sight (LOS) to Non-Line-of-Sight (NLOS) scenarios with beamforming and beamtracking by leveraging reflected and scattered radio waves [AT-1, INF-1]. Telecom vendors have demonstrated very high data rates using mmWave spectrum (e.g., 15 Gbps at 73 GHz with 2 GHz bandwidth [NOK-1]).

Likely NR Features

NR is still work-in-progress since its standardization is currently ongoing in 3GPP. However, it is possible to extract various likely NR features for example from the latest 3GPP documents, and vendor and operator white papers and press releases. These include:

• Access/backhaul integration [ERIC-1]
• Carrier Aggregation (CA) evolution [QC-1]
• Direct Device-to-Device (D2D) communication [ERIC-1]
• Flexible duplex (FDD (Frequency Division Duplexing) and TDD (Time Division Duplexing)) [ERIC-1] / dynamic uplink/downlink (UL/DL) switching [QC-1]
• Multi-antenna transmission [ERIC-1], which could include Massive MIMO (Multiple Input Multiple Output) [QC-1], FD-MIMO (Full-Dimension MIMO; 3D beamforming) [QC-1], Multi-User MIMO (MU-MIMO), hybrid beamforming, distributed MIMO, adaptive beamforming and beamtracking
• Multi-Connectivity (including NR, LTE, Wi-Fi) [QC-1]
• Narrowband 5G (NB-5G), supporting e.g., RSMA (Resource Spread Multiple Access) and multi-hop mesh [QC-1]
• NR-based MulteFire [QC-1]
• NR-based Licensed Assisted Access (LAA) [QC-1]
• NR-based tiered sharing of spectrum utilizing CBRS (Citizens Broadband Radio Service) and LSA (Licensed Shared Access) [QC-1]
• OFDM (Orthogonal Frequency Division Multiplexing) based waveform (also LTE uses OFDM) [3GPP-1]
• Scalable numerology with scaling of subcarrier spacing [QC-1]
• Scalable Transmission Time Interval (TTI) [QC-1]
• Self-backhauling [HW-1]
• Self-contained integrated subframe design [QC-1]
• Ultra-lean design [ERIC-1]
• User/control separation [ERIC-1]
• V2X (Vehicle-to-Everything) [QC-1, 3GPP-3]

It should be noted that many of these technologies are evolved versions of features (e.g., MIMO, carrier aggregation, Narrowband IoT (NB-IoT), MulteFire, LAA, LSA, OFDM) that have already been introduced in LTE. I intend to publish additional blog posts in the near future that give a bit more details about them.

5G beyond NR

Finally, even though they were the focus of this blog post, 5G is much more than a NR-based new radio access network and additional spectrum. There are plenty of developments also in the transport network and core network that deserve to be discussed in future blog posts, including for example network slicing, programmable network, service chaining (a.k.a., Service Function Chaining; SFC), Software-Defined Networking (SDN), Network Functions Virtualization (NFV), Next Generation Core (NGC), Cloud-RAN, virtual cells, xhaul, Core Network as-a-Service (CNaaS), RAN as-a-Service (RANaaS), end-to-end management and orchestration, Mobile Edge Computing (MEC), and the use of Machine Learning (ML) and Artificial Intelligence (AI). In addition to NR, many of these technologies will be needed to make 5G a platform that is capable of providing extreme bandwidth, extremely low latency, extreme reliability, extreme coverage, extreme flexibility, and security to cater for the needs of a broad range of use cases.

References

[3GPP-1] 3GPP TR 38.912 V1.0.0 (2017-03), 3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Study on New Radio (NR) Access Technology (Release 14), http://www.3gpp.org/ftp//Specs/archive/38_series/38.912/38912-100.zip

[3GPP-2] Tentative 3GPP timeline for 5G, http://www.3gpp.org/news-events/3gpp-news/1674-timeline_5g

[3GPP-3] 3GPP TR 22.886 V15.1.0 (2017-03), Technical Report, 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Study on enhancement of 3GPP Support for 5G V2X Services (Release 15), https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=3108

[AT-1] Millimeter-wave 5G modem coming mid-2018 with 5Gbps peak download, https://arstechnica.com/business/2016/10/qualcomm-5g-x50-modem-millimetre-wave-5g-modem/

[ERIC-1] Ericsson White paper, 5G Radio Access, https://www.ericsson.com/res/docs/whitepapers/wp-5g.pdf

[HW-1] 5G: A Technology Vision, http://www.huawei.com/5gwhitepaper/

[INF-1] Introduction to Millimeter Wave Wireless Communications, http://www.informit.com/articles/article.aspx?p=2249780

[ITU-R-1] Recommendation ITU-R M.2083-0 (09/2015), IMT Vision – Framework and overall objectives of the future development of IMT for 2020 and beyond, https://www.itu.int/dms_pubrec/itu-r/rec/m/R-REC-M.2083-0-201509-I!!PDF-E.pdf

[ITU-R-2] Recommendation ITU-R  P.2040-1 (07/2015), Effects of building materials and structures on radiowave propagation above about 100 MHz, https://www.itu.int/dms_pubrec/itu-r/rec/p/R-REC-P.2040-1-201507-I!!PDF-E.pdf

[LR-1] Vodafone CTO 'Worried' About 5G mmWave Hype, http://www.lightreading.com/mobile/spectrum/vodafone-cto-worried-about-5g-mmwave-hype/d/d-id/730679

[NGMN] NGMN 5G White Paper, https://www.ngmn.org/uploads/media/NGMN_5G_White_Paper_V1_0.pdf

[NOK-1] 5G Radio Access System Design Aspects, http://resources.alcatel-lucent.com/asset/200009

[QC-1] Making 5G NR A Reality, https://www.qualcomm.com/documents/making-5g-nr-reality

A Quick Guide to LTE in Unlicensed Spectrum, Part 2: SAS (Spectrum Access System) and LSA (Licensed Shared Access)

In my previous blog post, I covered a number of technologies related to the use of LTE in unlicensed spectrum. This blog post continues on that topic, with a focus on spectrum management systems.

Today, wireless spectrum below 6GHz is fully allocated to incumbents who, unsurprisingly, oppose any proposals to repurpose the spectrum to someone else. Thus, the traditional approach of repurposing spectrum for new cellular technologies is no longer sufficient in the case of 5G – novel ways to manage spectrum below 6GHz are needed. Two such technologies are LSA (Licensed Shared Access) and SAS (Spectrum Access System). LSA is a shared spectrum technology deployed in Europe, whereas SAS is its counterpart in the US.

Spectrum Access System (SAS)

SAS authorizes and manages use of spectrum for the Citizens Broadband Radio Service (CBRS), which is 150MHz of spectrum in the US in the 3550-3700MHz band that the FCC (Federal Communications Commission) opened up in 2015 for commercial use. CBRS enables almost anyone to use this spectrum while it is still being used by existing incumbents such as the military or satellite communication [QC-1].

CBRS defines three levels of priority when assigning shared spectrum [TSC]:

• Tier-1: Incumbent users (e.g., the Department of Defence) have the highest priority. 
• Tier-2: Priority Access Licenses (PAL), where an organization can pay for a fee to be assigned a 10MHz block for a limited geographic area for a three-year period, giving them priority over other users. A maximum of seven such blocks can be concurrently allocated within the same geographic area. Thus, max 70MHz of the available 150MHz can be assigned to priority users and the remaining 80MHz is left for others to use.
• Tier-3: General Authorized Access (GAA) for general usage. In this third tier, anyone can use the spectrum when it is not used by the higher layers.

As an example, an LTE network can be operated on licensed shared basis in the CBRS frequency band and it is not surprising that the telecom industry is interested in the band since it is considered one of the low-band options to support 5G, which is expected to have a high reliance on bands above the 25GHz level [RCR].

The 3.5GHz band does not penetrate indoors as much as lower frequencies. Thus, it is a good fit for small cells and their dense deployment model. As an example, organizations such as private enterprises, venues, and fixed operators could autonomously deploy high-quality in-building LTE networks e.g., following a floor-by-floor deployment model [FW].

Licensed Shared Access (LSA)

The LSA system includes coordination of resource usage between incumbents and LSA licensees. Its main use case is the extension of cellular capacity below 6GHz in Europe. One major difference between SAS and LSA is that LSA is missing tier-3 (i.e., GAA). LSA enables an LTE network to be operated on licensed shared basis in the 2.3-2.4GHz frequency band between incumbents (tier-1) and Mobile Network Operators (MNOs; tier-2) [Intel]. Tier-1 is prioritized over tier-2, meaning that the MNO using the LSA band is required to vacate it for a given geographic area, frequency range, and period of time for which the incumbent is requiring access. The expectation is that MNOs will enter in multi-year (typically ten years or more) sharing agreements with the incumbents for the LSA band. The LSA band is typically combined with LTE in dedicated licensed spectrum through carrier aggregation mechanisms

Spectrum management in LSA relies on an LSA Repository, which is a centralized database providing information about the availability of the LSA spectrum to multiple MNO networks [Intel]. Incumbents need to provide a priori usage information to the database on the LSA band’s availability. Based on this information, the LTE system is either granted access or requested to vacate the concerned bands using control mechanisms of an LSA Controller located within the MNO network. The LSA Controllers and the LSA Repository interact through an ETSI (European Telecommunications Standards Institute) defined interface. In LSA, no sensing mechanisms are required for the identification of incumbent operation in the LSA band.

And back to SAS

In contrast to LSA, SAS is designed to ensure coexistence with incumbents (e.g., the military using networks operated mainly close to US coastal areas) who are not able to provide a priori information to the central repository [Intel]. SAS can be operated throughout the US territory except within exclusion zones close to the coastal areas. In the second step of SAS, an Environmental Sensing Capability (ESC) component is added which allows operation even within the exclusion zones. ESC performs required sensing tasks to discover incumbent operation in the CBRS band. Spectrum access for tier-2 and tier-3 users is based on the sensing results. ESC will consist of networks of sensors that detect the presence of signals from incumbent systems in the CBRS band and communicate that information to one or more SAS repositories to facilitate protection of incumbent operations in the band [FW-2].

One of the companies that has filed SAS Administrator and ESC Operator applications with the FCC is Google [FW-2]. Google wrote in its application that it has been trialing a prototype SAS for almost two years. Google has also demonstrated its SAS prototype to industry and government. Google believes that 5G deployment will receive a boost from the CBRS band and that SAS will make the traditional model of MNOs spending billions of dollars to gain control of specific spectrum blocks no longer valid [RCR]. According to Preston Marshall, an engineering director for Alphabet Access at Google, “This [SAS] changes what spectrum ownership means. There is no owner”. Google believes that the killer application for the CBRS band will be the neutral host concept, which is an attractive model in places where it is not feasible for every operator to deploy their own radio systems independently, such as in enterprises or public venues like sports stadiums or shopping malls.

Besides neutral host scenarios, the CBRS band can be useful for enabling MNOs to offer Gigabit LTE speeds in more places thanks to the additional spectrum, for small cell deployments to extend coverage and capacity outdoors, and for creating private LTE networks for enterprises or industrial IoT.

Google is not stopping at the CBRS band – the company has urged the FCC to seriously consider applying the SAS framework to the 24GHz band as well [SE].

References

[FW] Google, Intel, Nokia and more partner to advance U.S. 3.5 GHz CBRS, http://www.fiercewireless.com/tech/google-intel-nokia-and-more-partner-to-advance-u-s-3-5-ghz-cbrs

[FW-2] Google, Federated Wireless, others apply to fill role of SAS, ESC for 3.5 GHz, http://www.fiercewireless.com/tech/google-federated-wireless-others-apply-to-fill-role-sas-esc-for-3-5-ghz

[Intel] Spectrum Sharing Technology | LSA and SAS White Paper, http://www.intel.com/content/www/us/en/wireless-network/spectrum-sharing-lsa-sas-paper.html

[QC-1] A new kind of spectrum for new opportunities, https://www.qualcomm.com/news/onq/2016/08/29/new-kind-spectrum-new-opportunities

[RCR] Google sees CBRS spectrum band as key for 5G, new model for industry, http://www.rcrwireless.com/20161117/carriers/google-sees-cbrs-spectrum-band-key-5g-new-model-industry-tag2

[SE] Google urges FCC to consider SAS model for 24 GHz band, http://www.spectrumeffect.com/google-urges-fcc-to-consider-sas-model-for-24-ghz-band.html

[TSC] What is CBRS Shared Spectrum for in-building small cell wireless? https://www.thinksmallcell.com/LTE/what-is-cbrs-shared-spectrum-for-in-building-small-cell-wireless.html

A Quick Guide to LTE in Unlicensed Spectrum, Part 1: LTE-U, LAA, LWA, MulteFire, Neutral Hosts, Private LTE Networks, QUIC and Multipath TCP

I have been a bit lost with all the unlicensed spectrum related terms and abbreviations listed in the title of this blog post. This post is an attempt to bring some clarity to myself and hopefully also to others who might discover the post.

Many of the technologies in the title like LTE-U, LAA, MulteFire and LWA are the result of recent developments within various standardization bodies and the telecommunications industry around data offloading to unlicensed spectrum.

LAA (Licensed Assisted Access) uses carrier aggregation in the downlink to combine LTE (Long Term Evolution) in unlicensed spectrum (5GHz, which is also used by 802.11a Wi-Fi) with LTE in the licensed band [QC-3, TVT]. The benefit of doing so is a fatter data pipe with faster data rates and more responsive user experience. LAA maintains a persistent anchor in the licensed spectrum. All control signaling information is carried in the licensed band, which ensures seamless and reliable user experience. LAA is for mobile operator deployments in Europe and Japan.

LTE-U (LTE in Unlicensed Spectrum) is the US (and Korean and Indian) version of LAA [4GA, QC-2]. It differs from LAA mostly in the way it handles contention. Whereas LAA has been specified in 3GPP, LTE-U has been defined in LTE-U Forum. LTE-U focuses on deployment options and regions where LBT (Listen Before Talk) policy, which is used by LAA, is not required. Both LTE-U and LAA dynamically select the unused channel with the least interference, avoiding Wi-Fi. At low traffic loads, they turn off transmission in the unlicensed spectrum, relying solely on the anchor of the licensed spectrum. In cases where there is no clear unused channel, they aim for fair and efficient coexistence with Wi-Fi. For this, LAA uses LBT to sense channel availability and adjust on/off LTE cycling (LTE on/off cycling is used to share the channel fairly with Wi-Fi). In contrast, LTE-U uses CSAT (Carrier-Sensing Adaptive Transmission) to sense other users and adjust on/off LTE cycling. CSAT and LBT represent fundamentally different approaches to access the channel [IEEE]. There are opinions that LTE-U/LAA are an attempt by mobile network operators to expand into the Wi-Fi spectrum and push out the Wi-Fi operators [TVT].

MulteFire is an LTE-based technology that, unlike LAA and LTE-U, solely operates in unlicensed spectrum without requiring an anchor in licensed spectrum [QC-1]. It broadens the LTE ecosystem to entities that may not own licensed spectrum, including ISPs (Internet Service Providers) and enterprise and venue owners. MulteFire is claimed to combine the performance benefits of LTE technology such as enhanced capacity, range, mobility and quality of experience with the simplicity of Wi-Fi like deployments. MulteFire is suitable for neutral host services. Neutral host is an LTE deployment that can be used by subscribers irrespective of their service provider. In the neutral host model, any deployment (e.g., by an ISP or enterprise or venue owner) can provide wireless access services to any end user without requiring a SIM or subscription, using neutral unlicensed spectrum. MulteFire can also interface with mobile networks to offer data offload services to mobile operators. Finally, MulteFire can be used to create private LTE networks using unlicensed spectrum bands [QC-4] for example for Industrial Internet of Things (IIoT) or enterprise use.

LWA (LTE Wi-Fi Aggregation) differs from LAA and LTE-U since LWA access in the unlicensed spectrum is based on IEEE 802.11 (i.e., Wi-Fi) standards [NOK]. LWA uses both LTE and WLAN (Wireless Local Area Network) interfaces in parallel: LTE provides reliable connectivity and mobility and Wi-Fi boosts data capacity and improves coverage. LWA aggregates data at the radio access network where Evolved NodeB (eNB; a base station) schedules packets to be served on LTE and Wi-Fi radio links [4GA].  LWA supports downlink aggregation at the PDCP (Packet Data Convergence Protocol; one of the user plane protocols in LTE that sends and receives packets to and from User Equipment (UE) and eNodeB over air interface) layer. It supports uplink transmissions only on the LTE network. LWA introduces a new direct interface Xw, which is defined between LTE and Wi-Fi. The Xw interface is terminated at the WT (WLAN termination), which is a new 3GPP logical node that may be in control of one or more WLAN APs (Access Points).

QUIC (Quick User Datagram Protocol Internet Connections) and Multipath TCP (Transmission Control Protocol) are options for multi-connectivity/aggregation that occur at the network level [4GA]. They can work with any combination of licensed and unlicensed band technologies, including future ones such as 5G. QUIC is a transport protocol that supports a set of multiplexed connections between two endpoints over UDP (User Datagram Protocol). It provides security equivalent to TLS/SSL (Transport Layer Security and Secure Sockets Layer) together with reduced connection and transport latency, and bandwidth estimation in each direction to avoid congestion [WP]. IETF’s (Internet Engineering Task Force) QUIC Working Group (WG) is looking into extending QUIC’s core protocol facilities to enable multipath capabilities for connection migration between paths and load sharing across multiple paths [IETF]. This would also enable the parallel use of Wi-Fi and LTE interfaces to download content. Multipath TCP is another ongoing effort in the IETF that aims at allowing a TCP connection to use multiple paths to maximize resource usage and increase redundancy [WP-2]. The use of both Wi-Fi and a mobile network is a typical use case for Multipath TCP. Multipath TCP can split data and transmit them through both LTE and Wi-Fi networks in parallel. LTE-Wi-Fi aggregation takes place between a device and MPTCP proxy, meaning that no dedicated network equipment is required for aggregation additionally [NM].

There exists a bunch of other exciting unlicensed spectrum related abbreviations and technologies that I plan to look into in a future blog post, including CBRS, LAS, SAS, PAL, GAA, and ESC. Stay tuned!

Although this blog post focused on LTE in unlicensed spectrum, unlicensed spectrum will also be key in boosting data rates in 5G NR (5G New Radio), which is the topic of another blog post.

(If I counted right, I managed to fit over 30 different abbreviations into this blog post).

References

[4GA] LTE Aggregation & Unlicensed Spectrum, http://www.4gamericas.org/files/1214/4648/2397/4G_Americas_LTE_Aggregation__Unlicensed_Spectrum_White_Paper_-_November_2015.pdf

[IEEE] Unlicensed LTE/WiFi coexistence: Is LBT inherently fairer than CSAT?, http://ieeexplore.ieee.org/document/7510910/

[IETF] IETF’s QUIC Working Group Charter, https://datatracker.ietf.org/wg/quic/charter/

[MG] Google’s QUIC protocol: moving the web from TCP to UDP, https://ma.ttias.be/googles-quic-protocol-moving-web-tcp-udp/

[NM] Analysis of LTE – WiFi Aggregation Solutions, http://www.netmanias.com/en/post/reports/8532/laa-lte-lte-u-lwa-mptcp-wi-fi/analysis-of-lte-wifi-aggregation-solutions

[NOK] Unlicensed band opportunities for mobile broadband, http://resources.alcatel-lucent.com/asset/200296

[QC-1] Introducing MulteFire: LTE-like performance with Wi-Fi-like simplicity, https://www.qualcomm.com/news/onq/2015/06/11/introducing-multefire-lte-performance-wi-fi-simplicity

[QC-2] How different are LTE-U and LAA?, https://www.qualcomm.com/news/onq/2015/11/04/how-different-are-lte-u-and-laa

[QC-3] Extending LTE to unlicensed spectrum globally – LAA, https://www.qualcomm.com/invention/technologies/lte/laa

[QC-4] GE, Nokia and Qualcomm Unveil First Private LTE-based Trial Network Customized for Industrial IoT, https://www.qualcomm.com/news/releases/2017/02/22/ge-nokia-and-qualcomm-unveil-first-private-lte-based-trial-network

[TVT] LWA offers aggregation without the aggravation: the alternative acronym to LTE-U and LAA, http://www.telecomtv.com/articles/wi-fi/lwa-offers-aggregation-without-the-aggravation-the-alternative-acronym-to-lte-u-and-laa-12804/

[WP] QUIC, https://en.wikipedia.org/wiki/QUIC

[WP-2] Multipath TCP, https://en.wikipedia.org/wiki/Multipath_TCP