Wednesday, September 27, 2017

"LTE for Layman (Part 2) - Scaling the LTE!"

Link: https://www.linkedin.com/pulse/lte-layman-scaling-aayush-shrut/

In my previous entry, we discussed at length about the various concepts and terminologies associated with the wonderful world of LTE. We learned about the history, technology behind radio towers and various components associated in establishing a LTE network. In this entry, we try to move one-step further in simplifying the complexities of LTE. Objective of this article is to help understand the protocol stack of LTE, how signal measurements are done, and the different channels associated with LTE.  Knowledge about LTE architecture is mandatory to understand the text here, so do go through the previous part in case you are unfamiliar with LTE terminologies. As always, focus is on simplicity, creating understanding, and demystifying the complex terminologies associated with LTE.  

The Yummy multi-layered Cake of LTE!

Imagine a multi layered cake that you generally get for wedding. Imagine the various flavours that it brings. Starting with vanilla as top layer, you get delicious chocolate layer, followed by strawberry, raspberry and butter scotch! Each flavour is unique in its taste, but only when combined together they form the perfect wedding cake. Each layer has its own role, to give a distinct flavour to it. If you remove any layer, the cake would taste quite differently and guests may even start complaining over lack of their favourite flavour! Only when the savouring taste of every layer is combined together, we would get happy guests and a delicious cake. This cake, in LTE terms, is fancifully known as LTE protocol stack!
Consider the protocol stack as complex set of rules, layered over one another just like a wedding cake. Each layer has its own function and own protocols (rules). While layer 1 would deal with actual physical constraints such as frequency considerations with its own set of protocols, layer 4 would deal with higher level functions such as which signal to transmit with their entirely different set of protocols. What makes protocol stack more special is the fact that each layer is inter-operable, or both vertically and horizontally aligned. This means that protocols for a particular layer on the sender side would be dealt with the same layer on the receiving side as well. This enables network operations to be done in an efficient way, with the combined protocol stack essential for complete coding-decoding of data from sender to receiver. Such layered protocols are common in any network consideration, with primary example being OSI or TCP/IP model.
Coming back to LTE protocol stack, they are divided into two separate stacks. One is for control-plane signalling communication between UE and MME, another one is for user-plane data communication between UE and PGW. 
                                        Control Plane Protocol Stack
                                           Data Plane Protocol Stack
The blue area is the one concerning the Access Stratum protocols, or the set of rules that dictate transfer of message between UE and eNodeB. The green area is the one concerning the Non Access Stratum protocols. These are set of rules that dictate transfer of messages between UE and core components of LTE network such as SGW, PGW and MME. The complete list of functionalities of each layer is beyond the scope of this article, however, a brief intro of LTE layers is mandated to understand the basic functioning of LTE protocol stack. These layers are unique for LTE. Other protocols such as GTP-U, SCTP and UDP/IP are globally used protocols whose intricate knowledge is not required to understand how your mobile connects. 
  1. L1: responsible for power control, measurement of air interface and initial search of LTE network. 
  2. Medium Access Layer (MAC): responsible for multiplexing of different resources, priority allocation of UEs through dynamic scheduling (de facto feature of LTE) and error correction through HARQ. 
  3. Radio Link Control (RLC): responsible for reassembly of fragmented packets and flow control (to handle congestion in network).
  4. Radio Resource Control (RRC): only found in Control Plane, this layer is responsible for everything signalling. This includes functions such as broadcasting of system information, paging, and establishing a constant radio bearer connection between different nodes. RRC is an important part of LTE network, as such; every UE must establish a RRC connection before communicating any data.
  5. Packet Data Convergence Protocol (PDCP): Sequencing, actual transfer of data to control or user plane, duplicate data detection, maintenance of sequence numbers of each packet, ciphering/deciphering, integrity protection/verification etc., these all are part of PDCP layer functioning. 
  6. Non-Access Stratum (NAS): It is primarily responsible for maintenance of IP address connectivity with the moving UE, along with host of other session management procedures.  
  7. Relay: More like an advanced repeater, it is essentially a mini eNodeB on top of eNodeB! It carries a subset of eNodeB functions, does the basic decoding of data, and is primary connected with another donor eNodeB farther away. The main use of relaying comes when your ever-moving mobile goes to the very end of tower range. It would than need support from "relays" of other towers to continuously buffer your HD YouTube video! 

Measurements in LTE:

Everything in the world that can be measured has a basic unit of measurement. You have cells, which combine to form tissues, which further combine to form organs. Matter has atoms, which combine to form, well, the universe! In wireless communication systems like LTE, it is bits, which combine to form bytes, which combine to form kilobytes and so on. However, it gets a bit complicated, as LTE in essence is an electromagnetic wave! There are two dimensions to consider in LTE signalling, each with their own basic unit. In horizontal axis, representation of a LTE wireless signal is done by time. In vertical axis, it is represented by frequency. Both of these dimensions combine to give the exact location of LTE signal, which is carrying, the bits of information.
The main unit of LTE used in horizontal or time axis is a frame, represented by 10 ms in length. Each frame is divided into 10 subframes, 1 ms in length. Each subframe is further divided into "time slots", 0.5 ms in length. Each of these time slots carry six (or seven based on certain criteria) OFDM symbols, 66.7 us in length. Do recall that LTE communication happens using OFDM, and therefore uses OFDM symbols to represent time domain. Consider these OFDM symbols as the lowest element of an atom that is a frame. A pictorial representation of a FDD frame is given below:
There is another representation of a frame used in Time Division Duplexing method (a type of two-way communication), in which a frame of 10 ms is divided into two half frames of 5ms. However, Frequency Division Duplexing is most widely deployed mechanism in cellular technology since the times of GSM, and the frame structure used throughout in most of the LTE texts is of that.
In vertical or frequency axis, the representation of LTE signal is done by "sub carriers", 15 kHz in height. No further division is needed here thankfully! Now, eNodeBs allocate resources or bits to the UEs in terms of "Resource Blocks" or RBs. These RBs are actual representation of "resources" used in LTE signal, carrying your bits of information! Just like water is a resource allocated by Government based on their criteria of demand and scarcity, eNodeBs allocate "RBs" or resources to your mobile (and countless others) based on their own algorithm of demand and scarcity. In LTE signal terms, think of "Resource Block" as the actual 2D grid representation of signal, represented by both frequency and time, carrying the information in the form of binary bits.
RB's actual dimensions are 12 points of vertical frequency axis or subcarriers by one point of horizontal time axis or time slot. In other words, 12 sub carriers in frequency by six OFDM symbols in time comprise one Resource Block, the "water" of LTE! One subcarrier by one OFDM symbol is called "Resource Element", the H2O of water if analogy is drawn. These "Resource Elements" represent "Modulation Symbol", which carry one to six bits of data depending on the modulation scheme used. For example, a QPSK modulation scheme would mean 4 bits of data is transferred in one resource element. A pictorial representation is given below:
Now, given the bandwidth is 10 MHz, or in other words, you are given 10 MHz of frequency or vertical axis. How many resource blocks can you form from that? Recall that one sub carrier is 15 kHz in length (or height). Therefore, for 10 MHz, it would be 10 MHz/15 kHz = 667 sub carriers. However, in LTE, there is a concept of "usable subcarriers". In other words, some subcarriers are used as "guard" to protect against noise and distortion. According to release 8 of LTE specifications, 33 subcarriers are assigned in both front and back end of a LTE signal to act as guard. In addition, one subcarrier acts as a centre point in which no information is sent, otherwise called DC subcarrier. It is an important subcarrier in OFDM systems, as new mobiles trying to connect to LTE system for the first time uses the DC subcarrier to locate centre of LTE frequency band (discussed later). Therefore, this makes total of 33*2 + 1 subcarriers unusable for information transmission. This leaves 667 - 66 - 1 = 600 subcarriers usable for data transmission. Now, one resource block is 12 subcarriers. So, 600/12 = 50 resource blocks can be assigned in 10 MHz of frequency or vertical axis assigned. 

Transportation in LTE:

So far, we have discussed about various layers of LTE cake, and how are resources allocated in frequency/time domain by eNodeB. Now without a mechanism to carry these resources over all these layers, the information would lie stagnant! Just like your water is supplied by government in pipes, RBs are allocated by eNodeBs and transported through lower layers by something called "channels". 
Think of channels as conveyor belts used in factories. So one belt or channel processes your data and signal in a different way, than passes it along to another belt or channel and so on. LTE system uses three different channels, each being primarily used to segregate information processing. The three channels are:
  1. Logical Channel: Used for communication between RLC - MAC layer interface. Denotes "what type" of information is carried, such as "broadcast message" or "control signal message" etc. 
  2. Transport Channel: Used for communication between MAC - PHY layer interface. Denotes "how is" information carried, such as encoding options, error correction mechanism, modulation scheme etc.
  3. Physical Channel: Used for communication between PHY - Air interface. Denotes "where is" information carried, the exact location of bits in a LTE signal.
Most of the channels are mapped onto each other in following fashion, based on uplink and downlink direction:
While the acronym explanation of each channel would expand this already lengthy article, an overview of critical physical channels is warranted. This is because every channel is at last mapped to Physical channel, the last conveyor belt in the cake factory of LTE. Think of Physical Channel as both provider of coordinates of your data, and the transporter of them over the air!
  1. PDCCH/PUCCH stands for Physical Downlink/Uplink Control Channel: It is used for conveying scheduling allocation information for data or RBs. These allocation information are very important, as they give functioning parameters to your mobile such as modulation scheme to be used, scheduled RBs for uplink (or where to put your #instagram selfie for upload in frequency/time grid), downlink data location etc. Essentially, the coordinates of data channels PDSCH/PUSCH are provided on this channel. This channel is mapped from something called Downlink Control Indicator (DCI) and Uplink Control Indicator (UCI) of Transport Channel. PDCCH is always mapped in first L OFDM symbols, given by PCFICH. PUCCH on the other hand occupies either ends of system bandwidth in terms of location, and size is given by something called System Information Blocks (SIB-2 in this case). SIBs carry important channel information, and are of different types. We would discuss more of this in the next part. For now, think of PDCCH/PUCCH as your investment banker, one who knows exactly where your money is and how to access them, or the one who provides you the resources to invest your money.
  2. PBCH stands for Physical Broadcast Channel – It carries the channel bandwidth information, antenna configuration and reference signal power that broadcast to every cell phone in the eNodeB coverage area. Think of this channel as advertiser of LTE band, one that constantly shouts to every cell phone about the critical system information needed to initiate first connection with LTE network, i.e. when mobile is switched on for the first time or when a new cell phone appears in the LTE network. eNodeB advertises this using something called “Master Information Block” (MIB), which carries the bandwidth information. 
  3. PDSCH/PUSCH stands for Physical Downlink/Uplink Shared Channel – It is a common, big channel used for carrying both data and signalling messages. Think of it as a double decker bus, carrying control signals, system configuration messages and data towards the destination. Its size is determined by reading DCIs from PDCCH.
  4. PCFICH stands for Physical Control Format Indicator Channel - It carries the number of OFDM symbols that can be used for control channels (PDCCH). It remains mapped in the first OFDM symbol. Sixteen resource elements are allocated for PCFICH, whose location is determined by cell id and bandwidth. Think of it as the location provider of your investment banker!
  5. PHICH stands for Physical Hybrid ARQ Indicator Channel - It carries the acknowledge signal of properly receiving and decoding the uplink message sent by UE. Its size is determined by MIB, and location is on first OFDM symbol, same where PCFICH resides.
  6. PMCH stands for Physical Multicast Channel - This channel is used for transmission using multicast technique, or transmitting information to some selected UEs simultaneously. 
  7. PRACH stands for Physical Random Access Channel - When the UE wants to make the initial request for access to LTE network, it uses this channel. As both location and identity of UE is unknown for core LTE network, PRACH is the only channel that can be used by UE for non-synchronized access to the network. Six resource blocks are allocated for this channel, with starting location communicated in SIB.
Given below is a diagram of critical uplink and downlink channel, as to how they are allocated in frequency/time grid:
Apart from the channels, there is a concept of "Reference Signals" and "Synchronization signals" in LTE. These signals are something that are unique to Physical Layer, and are not aligned with other channels or higher layers. 
1.) Reference signals: They are used by UE to determine power level information of the downlink channel. UE uses this power level information as a "reference" to take several measurements, which aids in determining channel conditions, handover scenario and even cell selection. There are different types of Reference signals whose location in frequency/time axis is based upon the antenna ports used for transmission of these signals. Given below is a downlink reference signal (one antenna configuration), where resource element is allocated in first and third to last OFDM symbol.
2.) Synchronization Signals: They are used by UE to have slot level and frame level synchronization with the LTE network, the very first time your UE boots up. Synchronization signals are bandwidth independent, and the UE starts to tune into sync signals from smallest frequency of 1.4 MHZ. Moreover, for every frequency, they have fixed position, occupying time slot 0th and 10th around the central slot (determined by DC subcarrier) of LTE frame. In other words, the X-axis coordinate (time location) of the first signal that the UE acquires is always fixed in every frequency station (Y-axis) that the mobile tunes in. There are two types of Synchronization signal: primary sync signal that occupies the last OFDM symbol and secondary sync signal that occupies the second last OFDM symbol. In total, 6 RBs or 72 subcarriers are allocated which are centred on the DC subcarrier, however only 62 subcarriers are actually used. Synchronization signal is the very first signal that UE acquires, and is used by UE to determine frequency/time grid of eNodeB and the Physical Cell Identity (PCI). Given below is a diagram of Sync signal:
If all of this sounds confusing (it sure is!), here is a step by step logical way of deciphering channel access:
  1. Your mobile, when turned on for the first time, acquires sync signal using DC subcarrier (center of any frequency band).
  2. From sync signals, UE derives physical cell identity.
  3. It than acquires the bandwidth using MIB broadcasted in PBCH, again using help of DC subcarrier.
  4. From cell identity and bandwidth acquired earlier, it derives location of PHICH.
  5. On Reading PHICH, UE knows the OFDM symbols for PDCCH that holds control location information stored in DCI.
  6. Reading the DCIs, UE acquires location of critical SIBs stored in PDSCH.
  7. From the information gathered after reading SIBs, UE is ready to make the initial connection using PRACH channel.
  8. After connection is established, UE transfers your data using PUSCH channel and receives on PDSCH channel.
Bottom line is that everything associated with a wireless LTE signal is located in the frequency/time grid. It is just that channels provide a means to segregate the information, and locate every information coherently!  

Conclusion: 

This concludes second part of LTE for Layman series. We have discussed how entire LTE is just a series of protocols separated from one another by set of layers. We also got an overview of LTE frame structure and sub carriers, on how LTE does its measurements in frequency/time axis and forms resource blocks using it. We than dived into the world of channels, on how they are allocated and what do they carry. Now armed with this knowledge, we are ready to understand the entire lifecycle of UE, from first signal acquisition to finally uploading your #instagram selfie in next part. Do like and share your opinions, as wise men once said: “Sharing is caring”!

"LTE for Layman (Part 1) - Introduction and Architecture" by Aayush Shrut

What is 4G and how does it actually work? Internet is filled with quality scholarly articles, and you can find numerous books explaining the various intricacies of LTE (Long Term Evolution) in detail. My personal favorite is the classic: "LTE: The UMTS Evolution - From Theory to Practice" published by Wiley. However, sometimes the complexity and terminologies get too much over our head in establishing a bare minimum idea, particularly to those who are not from technical background. This series is a humble attempt in trying to remove all the complexities associated with technology. Using power of analogy with a sprinkle of humour, readers can get a better understanding of the otherwise vast 4G technology, filled with complexity and constant developments! Objective of this article is to give readers a bird's eye view of LTE: origin from 3GPP, the underlying electronics, and the architecture components.

A bit of history first!

Everything that your mobile does, the kitten videos you watch on youtube, the calls you place, the call drops that happen, or those awesome selfies that you share on #instagram, everything is governed by a set of rules or protocols (some say, even life). These set of rules (analogous to code) decide how to process the selfie, when you press on the upload button. It includes the complete conversion of data into digital format, travelling over the air through towers, going through some complex processing and then eventually finding itself on the internet. These rules of processing are governed by 3GPP: 3rd Generation Partnership Project, which is essentially an association of major telecommunication giants who work tirelessly in researching ways and means to create faster mobile internet speed! The 3GPP organizational partner releases a set of rules/protocols after meetings conducted with almost all major telecommunication players. These protocols and the associated architecture are considered as worldwide standard, which is followed by all LTE stakeholders across the globe. Why do we need 3GPP? Well, if it had not been for them, than Airtel 4G Sims would work only on selected mobile such as the ones created by Apple only! In short, SIM cards would have had become Apple (end to end control anyone?).
The 3GPP people gave us 2G GSM, 3G UMTS, and now 4G LTE. All these terms, GSM, UMTS and LTE, they are essentially set of rules and standards developed by 3GPP, due to which we can communicate and exchange selfies worldwide! In addition, for historical context, LTE rules were released in 2008 under "release 8" of 3GPP specifications. There is also a set of advanced rules for LTE called 4G advanced, which is just an advancement on base release.

Electronics 101

"Behind every communication technology is a radio wave". This holds true for LTE, which is nothing but means to make the radio waves transfer faster and more efficient. The claimed speed of LTE is approx. 300 Mbps downlink, and 75 Mbps uplink based on certain antenna conditions (actual speed and coverage differs country to country). It also promises to reduce latency (or delay in data transfer) up to five ms. In order to uphold the promises; significant overhaul in the underlying radio transfer technology was done to enable the high-speed transfer.
While the entire listing of changes is beyond the scope of this article, one major change factors in a lot. It is called OFDM or Orthogonal Frequency Division Multiplexing. 
Let us start with the basics of signal communications, i.e. Modulation. It simply means that a radio wave is modified with another radio wave carrying information which than together, can be transferred over long distances. The combined radio wave is called "carrier" signal, which transmits the original signal over long distances using process of Modulation. What if we want to transmit different information, e.g. black + white, in the same "carrier" signal? We create "sub carriers", which are nothing but parts of the main carrier signal, carrying different levels of additional information. 
How to transmit these information on the limited bandwidth (or frequency channels) available? The traditional method (since 1870!) is by using Multiplexing, where multiple signals are clubbed together (or multiplexed) over a frequency channel. This is called Frequency Division Multiplexing (FDM). However, to make the traditional approach even faster, OFDM came into picture. In typical FDM, the total bandwidth is divided into several non-overlapping frequency bands, each of which is used to carry signal (see diagram below). Simplest application of this is in Radio, where we tune into different frequencies being broadcast at the same time. The extra O or "Orthogonality" is an advanced geometrical concept incorporated with FDM, which overlaps the different sub carriers "orthogonal" to each other, hence saving bandwidth and increasing transmission capacity. In Layman terms, OFDM helps various sub carriers, carrying different data, transmit in cost effective manner, saving bandwidth (by providing orthogonality) and enabling faster transmission which is required for LTE. 
                Orthogonality transmission of sub carriers saving bandwidth
Now, to transmit so many of these sub carriers (analogy of tuning into different radio stations simultaneously), it would require several receiving and transmitting stations, one for each sub carrier. To eliminate this, Inverse Fast Fourier Transformation comes into rescue, which essentially transforms different sub carrier signals of varying frequency and strength into one combined time varying signal! In Layman terms, take one signal, divide it into different sub carriers carrying different information, and combine it into one signal to eliminate usage of different antennas, than reverse the steps on the receiving side! A more pictorial representation is given below:

This is the very basics of LTE as far as electronics aspect is concerned. Please do note that there are many additional concepts related to digital signal processing of LTE. There is cyclic prefix, which is essentially sending out redundant information to prevent signal distortion in the forms of Inter Signal and Inter Carrier interference. There is Single Carrier FDMA (SC-FDMA), which uses an additional technology of Discrete Fourier Transform to eliminate Peak to Average Power Ratio (PAPR) (nonlinear distortion of radio waves combined because some waves have much higher magnitude). It is used extensively in uplink transmission, as vast number of cell phones generate generates their own independent radio waves, summation of which results in a distorted radio wave signal. To save us from further electronics, let us move onto the Information Systems part of LTE!

LTE architecture, components and interfaces

If the above diagram looks mystifying, let us approach it component-by-component, beginning from the left.
1.) UE, User Equipment: it is your mobile, or any device that allows 4G network communication. In other words, any equipment that adheres to the rules set by 3GPP, and communicates with the LTE towers, to allow your #tag tweets to reach the internet faster qualifies as UE. Examples include tablets, smartphones, dongles or Apple! It scans the frequencies when you turn it on, latches on the best LTE frequency available and the magic begins from there on. In addition, of course, it has the coveted SIM card (subscriber identification module)! This little SIM makes sure your mobile is authorized to access the network, be registered, and be billed for the services it access. UE has a host of other advanced features such as measuring channel conditions (or whether the weather is stormy or not) in the form of Channel Quality Indicator, providing Scheduling Request if it has a lot of data, and among other things trying to maintain constant connectivity with the radio tower. 
2.) E-UTRAN, evolved UMTS (Universal Mobile Telecommunication System) Terrestrial Radio Access Network: Sounds a mouthful right? Well, in essence, E-UTRAN is a network created by LTE radio towers otherwise called Evolved Node-B or eNodeB. These eNodeBs combine to form the heart of the powerful LTE network called E-UTRAN. It's the job of eNodeBs to control the otherwise dumb UEs, send them signalling messages to perform various tasks, handle their over the air signals, dynamically schedule message exchange flow based on algorithms, and forward core network signals called Non Access Stratum (NAS) messages. In the era of 3G, there were two nodes, with one being the radio network controller (RNC) and the other just a radio tower. These nodes were combined together to form eNodeB so as to enable LTE to do what it's advertised to do, reduce latency and increase the download speed! Otherwise called Access Network, E-UTRAN and the associated eNodeBs have no centralized controller hence having a flat architecture. They connect with the core network using an interface called S1, and through each other using an interface called X2. These interfaces, in analogical terms, are just fancy names for a gate to connect to outside world. Therefore, a gate, which connects you directly to your neighbour’s home, would be called X2, and the gate, which connects you to municipality headquarters, would be called S1. 
3.) EPC, Evolved Packet Core: This is the backend, the LTE subscriber's home, Airtel 4G woman’s office. EPC handles the creation of bearers, which is the heart of data transfer. Think of bearers as tunnels in the air, carrying all your data from the Internet to your mobile. This end-to-end connectivity management through the eNodeB is the work of EPC and all of its components. Other functions include billing for the data connection, providing IP address for your cell phone, authenticating your SIM card and providing fancy ciphering techniques to secure your data flow. An all IP network, EPC is part of System Architecture Evolution (SAE), the core network of LTE where every magic happens. The components of EPC, otherwise called SAE Core are:
a.) MME, Mobility Management Entity: The Mother of EPC and perhaps your cell phone, it handles the creation and destruction of the bearers that carry your data. In addition to that, it has host of other functionalities such as authentication, providing identities to UEs, selecting appropriate SGW and PGW nodes, tracking UE when it goes idle, and among other things, being the back end of all management related procedures of mobile. MME uses the functionalities of another node called HSS (Home Subscriber Server), which is essentially a comprehensive database of all the subscription and user identity information (the information you fill when getting a SIM). Analogy can be drawn in a way: if UE is the front end GUI of a software and E-UTRAN is the middleware for operations, MME + HSS would be the back end for database access, signalling and authorization. Since MME is the manager of eNodeB, the interface between MME and eNodeB is often called Control Plane. Analogy to how your manager controls you at work by sending signals such as "what is the status"?
b.) SGW, Serving Gateway: SGW is the router of your cell phone! As simple as it gets, it routes the data from the vastness of the internet towards your sophisticated E-UTRAN network using the bearers created by MME. Although part of EPC, SGW has a direct connectivity with eNodeB. As SGW main role comprises of forwarding user traffic from PGWs to the UE, the interface between SGW and eNodeB is appropriately termed as User Plane. Apart from the routing functions, it also acts as a mobility anchor for the handover of your data to other eNodeBs in case you move out of coverage from your parent eNodeB. Same anchoring functionality is provided by SGW if you move to other networks such as 2G/3G. Other functionalities include paging your cell phone in case it becomes idle and storing context information such as bearer details. 
c.) PGW, Packet Data Network - Getaway: If SGW can be considered as a router, PGW can be considered as a switch or a getaway (or gate to another world) to connect to the external network (otherwise called Packet Data Network). Think of PDN as external IP based network such as internet or IP Multimedia SubSystem (eg: Skype). Now think of assigning one PGW for each of these external PDNs. Unlike other nodes, where a UE can be connected at a time to only one eNodeB, one MME and one SGW, a UE can be connected to multiple PGWs based on the type of PDN. So for example, you can browse your internet and simultaneously have a VoIP call, as your UE would be connected to two PGWs, each connected to their respective PDNs. Each PGW would be acting as a getaway between the external and internal network, forwarding data to SGW who will feed it into bearers (terminating at UE's end) created by the MME! Other functionalities of PGW include maintaining the Quality of Service (QoS) for those Skype Calls and video surfing, and providing fair billing for the same. PGW does the QoS enforcement using help from another node called Policy and Charging Rules Function (PCRF). PCRF ensures that only the authorized user gets the desired quality of service, with the proper billing. So for example, if you have an HD video LTE pack, it is the job of PCRF to ensure you get HD videos. Hence, no one with a SD video pack will be able to access HD videos without the subscription.
This concludes the introduction and architecture overview of LTE. So what will happen if you like this post using your LTE enabled phone? Your UE will transmit information over the Uu interface to your parent eNodeB, who will than transmit the information to MME, who will create a bearer and select the appropriate SGW and PGW. The PDN would be selected as the internet, and the associated PGW will transfer information on the bearer created by MME via SGW, all this under 5 ms! Sharing is Caring!

Tuesday, September 26, 2017

AIPLA - In re Cray (after TC Heartland)

Patents/Venue
Federal Circuit Clarifies Venue Factors For
Regular and Established Place of Business 
The home office of a sales executive for a Washington corporation who worked for two years in the Eastern District of Texas was insufficient to establish a "regular and established place of business" in that district for venue under 28 U.S.C. 1400(b), the Federal Circuit held September 21, 2017. In re Cray, Inc., Fed. Cir., No. 2017-129, 9/21/2017.

The decision clarifies venue determinations based on a defendant's place of business after the Supreme Court recently rejected the Federal Circuit's views on determining residency-based venue. According to the appellate court, the crucial consideration in this case was that the employee's home did not have sufficient attributes of a place from which the defendant-employer's business was carried out. 
 
In addition, the district court incorrectly held that a defendant need not show a fixed physical location in the district to demonstrate proper venue, the Court concluded. While the "place" of business need not be a formal office or store, "there must still be a physical, geographical location in the district from which the business of the defendant is carried out." Merely showing that there was a physical location where an employee of the defendant carried on certain work for his employer was not enough here to establish that the employer maintained a regular and established place of business in the district. 

Background

After Raytheon Company sued Cray, Inc. in the Eastern District of Texas for patent infringement, Cray, a Washington corporation, alleged improper venue and moved to transfer the case to the Western District of Wisconsin. Raytheon conceded that residence-based venue in Texas is improper after TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S.Ct. 1514 (2017). However, it claimed venue was proper because Cray committed infringing acts in the Eastern District of Texas and had a regular and established place of business in that district.
 
The district court agreed with Raytheon, concluding that the activities of Cray's sales executive for two years in the Eastern District of Texas supported the infringement allegations and demonstrated that Cray had a regular and established place of business there. Cray petitioned the Federal Circuit for a writ of mandamus to transfer the case based on improper venue.

Federal Circuit Finds Venue Improper

The appellate court disagreed with the district court and granted Cray's the mandamus petition. According to the Federal Circuit, the district court misapplied the venue statute by failing to give sufficient weight to the literal language of Section 1400(b).  Judge Lourie acknowledged that the "regular and established place of business" language of the statute has only been addressed once in the context of a mandamus petition, citing In re Cordis, 769 F.2d 733 (Fed. Cir. 1985). However, the district court misunderstood the scope of Cordis and misapplied it to the facts of this case, the appellate court held. 
 
Judge Lourie acknowledged that business has changed since the 1983 Cordis decision, noting that the brick-and mortar model is not always the rule, that a business can be conducted virtually, that employees increasingly telecommute, and that retailers may not always warehouse products. However, TC Heartland requires a "focus on the full and unchanged language of the statute, as Cordis did not consider itself obliged to do," he wrote.

The 19th century predecessor of the patent venue statute clarified that infringement defendants could be sued outside of their state of incorporation, the Court explained. While it broadened venue, however, the provision did so by defining exact limits of patent venue, according to the Court. Judge Lourie said courts should be careful not to conflate showings meant for other purposes, e.g., personal jurisdiction or the general venue statute, with the showing required to establish proper venue in patent cases. The regular and established place of business standard requires more than the "minimum contacts" for personal jurisdiction or "doing business" for general venue, he added.

"Place of Business" Means "Physical Place"

Although the "regular and established place of business" requirement depends on the facts of the case, "the analysis must be closely tied to the language of the statute," Judge Lourie wrote. After consulting dictionaries contemporaneous with the 19th century enactment of the statute, he analyzed three requirements in the statutory language.

Place: This term requires a physical "place" in the district, and the district court erred in holding that a fixed physical location in the district is not a prerequisite. A virtual space is not enough. While the "place" need not be a formal office or store, there must still be a physical, geographical location in the district from which the business of the defendant is carried out.

Regular and established: The term "regular" refers to steady and uniform, not sporadic activity; while a single act would not be sufficient, a series of such acts would. "Established" means stable or permanent. While a business can move its location, it must for a meaningful time period be stable and established. If an employee can move his or her home out of the district at his or her own instigation and with no employer approval, the employee's home is less likely to be considered the employer's place of business.

Of the defendant: The place of business must be a place of the defendant, not solely a place of the defendant's employee. The defendant must establish or ratify the place of business; it is not enough for the employee to do so on his or her own.

Relevant considerations include whether the defendant owns or leases the place, or exercises other attributes of possession or control over the place. One can also recognize that a small business might operate from a home; if that is a place of business of the defendant, that can be a place of business satisfying the requirement of the statute.

Another consideration might be whether the defendant conditioned employment on an employee's continued residence in the district or the storing of materials at a place in the district so that they can be distributed or sold from that place. ... Marketing or advertisements also may be relevant, but only to the extent they indicate that the defendant itself holds out a place for its business.

In this case, Raytheon failed to prove that the home of the sales executive was the regular place of business of Cray, according to the Court.

The fact that Cray allowed its employees to work from the Eastern District of Texas is insufficient. There is no indication that Cray owns, leases, or rents any portion of [the sales executive's] home in the Eastern District of Texas. No evidence indicates that Cray played a part in selecting the place's location, stored inventory or conduced demonstrations there, or conditioned [his] employment or support on the maintenance of an Eastern District of Texas location. No evidence shows that Cray believed a location within the Eastern District of Texas to be important to the business performed, or that it had any intention to maintain some place business in that district in the event [the sales executive] decided to terminate their residences as a place where they conducted business. ...

For purposes of §1400(b), it is of no moment that an employee may permanently reside at a place or intend to conduct his or her business for present and future employers. "The statute clearly requires that venue be laid where 'the defendant has a regular and established place of business, not where the defendant's employee owns a home in which he carries on some of the work that he does for the defendant." [American Cyanamid Co. v. Nopco Chem. Co., 388 F.2d 818, 820 (4thCir. 1968)].

To read the Court's opinion in this case, click here.

[Law360] Applying Basic Antitrust Principles To SEPs

https://www.law360.com/ip/articles/967424/applying-basic-antitrust-principles-to-seps?nl_pk=b1903a90-09ca-421f-8c29-9e09cc40f173&utm_source=newsletter&utm_medium=email&utm_campaign=ip

Applying Basic Antitrust Principles To SEPs

Yee Wah Chin %>
Yee Wah Chin
The United States antitrust approach to intellectual property has evolved over time. The IP laws and the antitrust laws are now commonly viewed as complementary. Both value innovation, competition and consumer welfare, while IP rights are considered to be a form of personal property rights[1] that confer only the right to exclude others from the areas covered by the IP.[2] The same antitrust analysis applies to conduct involving IP as to conduct involving other forms of property, taking into account the specific characteristics of the particular property right.

There is no presumption that IP creates market power. The Patent Act makes that clear in the context of patent law.[3] The U.S. Supreme Court’s decision in Illinois Tool Works Inc. v. Independent Ink Inc.[4] extended that principle to the antitrust context.

However, there have been significant calls recently for findings that infringement suits and licensing conduct by patent assertion entities labeled “patent trolls” and holders of standard-essential patents generally are monopolization or attempts to monopolize that violate Sherman Act §2, 15 U.S.C. §2. This article argues that the basic principles of keeping in mind history and context, and general antitrust principles, apply equally to SEPs and PAEs as to other economic phenomena.

History and Context

The context of the state of a country’s economy affects its law, including competition law. Many of the early, key U.S. Supreme Court decisions interpreting the Sherman Act and laying the foundations of U.S. antitrust law involved huge cartels that affected substantial portions of the U.S. economy, such as the Standard Oil trust,[5] the railroad trust[6] and the meat packing cartel.[7] These are cases that may be unlikely to occur today in the U.S., substantially precisely because of this early law enforcement and because of changes in the U.S. economy in the last 120-plus years.[8]

Some of the evolution in the U.S. of the balance between fostering innovation and ensuring public access to innovation, and of the approach of antitrust law to IP, occurred perhaps as a result of the shift of the U.S. from being primarily an IP-taker in the 18th and 19th centuries, to significantly an IP-giver today. Some of this evolution can also be seen in other jurisdictions, perhaps including China.[9] These may be situations of where one stands depending on where one sits.

Antitrust Principles

The most common IP scenarios raising competition law concerns involve the unilateral actions of individual IP holders that may rise to monopolization and attempted monopolization.

The U.S. Supreme Court established in U.S. v. Grinnell Corp.[10] that monopolization requires two elements: first, that there is monopoly power in a relevant market, and second, that the power was acquired or maintained willfully as distinct from growing or developing as a result of a superior product, business acumen or historic accident. In NYNEX Corp. v. Discon,[11] the Supreme Court clarified that for the second element, simply having an anti-competitive motive, even if there was fraud that enabled a monopolist to raise prices, is insufficient to constitute a violation of the Sherman Act if there was no harm to the competitive process as a result.

Where there may be insufficient market power for monopolization, there may be attempted monopolization if there are: (1) the specific intent to destroy competition or achieve monopoly; (2) some exclusionary or anti-competitive conduct pursuant to that intent; and (3) a dangerous probability of success.[12]

Application to IP Generally

The U.S. antitrust enforcement agencies have consistently adhered to these basic antitrust principles for IP. This may be seen most recently in the 2017 update[13] to their 1995 "Antitrust Guidelines for Licensing of IP," and in the Federal Trade Commission’s 2016 report on PAE activity.[14] The 2017 update to the IP guidelines reaffirm the agencies’ enforcement approach with respect to IP licensing and do not expand the guidelines beyond licensing. The update is intended to conform the 1995 IP guidelines to changes in statutory and case law since 1995.[15]

In its PAE study[16] the FTC suggested steps that may address some concerns relating to PAEs, which are consistent with the principle stated in its October 2003 report “To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy,” and reiterated in its March 2011 report, “The Evolving IP Marketplace: Aligning Patent Notice and Remedies with Competition,” that where concerns with how patent rights are being exercised are based on aspects of the patent system that allow such abuses, the remedy may be more appropriately in adjustments in the patent system, unless the abuses clearly have an adverse impact on competition.

Antitrust Approach to SEPs

One definition of standards is “a set of characteristics or qualities that describes features of a product, process, service, interface or material.”[17] The process of standard-setting is that of identifying, developing and/or choosing such a set of characteristics or qualities. Standard-setting has beneficial effects, increasing consumer welfare and efficiency by establishing uniform approaches that enable interoperability and scale. However, the standards-development process, and standards themselves, may be abused and create anti-competitive effects.[18]

Many standards incorporate IP and require licenses of IP to be implemented. Under many standard development organization policies, a patent is “essential” to a standard if it is not possible as a technical matter to implement the standard without infringing the patent. There is a significant concern with SEP holders controlling and misusing market power that they obtained only because their IP is essential to implement a standard. There may be a failure to disclose SEPs during the standard development process, or a refusal to license SEPs, or refusal to license SEPs on fair, reasonable and nondiscriminatory terms. While all the cases brought by the U.S. agencies involving such allegations have been resolved by consent decree or ultimately dismissed,[19] and sometimes under the unfair practices prong of the FTC Act instead of the unfair methods of competition prong,[20] the competition law approach of the U.S. agencies to standards remains adherence to basic principles.

There are at least two levels at which the antitrust analysis should be made. First, there is the question of whether the particular standard has market power. If the standard is not dominant, does not have the power to control prices or exclude competition, then it would appear difficult to argue that any IP that reads on the standard, whether or not a SEP, has market power by virtue of reading on the standard.

Second, even if the standard involved has market power, there is the question as to whether any particular IP that reads on the standard has market power. In fact, where a SEP is subject to contractual commitments to license the SEPs under RAND terms, the SEP holder substantially limits the circumstances in which it may refuse to grant a license or obtain injunctive relief against infringement, and the ability of the patent holder to raise prices or exclude competition may be constrained.

There is debate in the U.S. as to whether the “nondiscriminatory” aspect of contractual RAND obligations would be best enforced as a contractual matter or whether competition law remedies are required.[21] The adjudicated cases in the U.S. relating to RAND terms have been contract law and patent infringement cases.[22]

Antitrust Approach to PAEs

PAEs hold patents, but do not practice them, and gain revenues primarily from enforcing the patents against other entities. The following three aspects of the FTC’s PAE study are striking:

1. The FTC took the time and effort to gather facts before taking or recommending any action.

2. The FTC recommended several changes in procedures for patent litigation, and is silent on any application of the antitrust laws.

3. The recommendations are apparently based on a view that while there appears to be a significant amount of nuisance infringement lawsuits brought by a significant percentage of PAEs, those suits have little impact on competition, so that the appropriate remedy for such abuse of process should be adjustments to the process to make it more difficult to abuse.

Infringement Suits by SEP Holders and PAEs

Patent infringement suits may be used in an abusive manner by patent holders, including SEP holders and PAEs. Nonetheless, under the Noerr-Pennington doctrine,[23] lawsuits may be brought for anti-competitive purposes, unless the lawsuit was shown to be objectively baseless under the standard established in Professional Real Estate Investors v. Columbia Pictures Industries.[24] That case establishes a 2-pronged test for whether there is a sham litigation that is unprotected by the First Amendment:

1. Whether the lawsuit is objectively baseless; and

2. If it was objectively baseless, whether it was brought without caring whether it will be won, because the goal is to affect the competitor by forcing the competitor to defend the lawsuit.

Implementing a criterion of “objectively baseless” for lawsuits brought by a PAE, that is based on a reasonable litigant’s expectation of success on the facts and law applicable to a particular case, would serve a dual purpose — discouraging the abuse of litigation and governmental process for anti-competitive purposes while preserving the proper use of patent infringement suits for enforcement of patent rights. Nonetheless, even if a lawsuit is a sham and brought for anti-competitive purposes, for the lawsuit to be an antitrust violation, the elements of monopolization or attempted monopolization must still be established.

Conclusion

Competition law is not a hammer to which all problems are nails. Sometimes, other remedies are more appropriate. The agencies' 2017 IP guidelines and the FTC’s PAE study are two examples of the adherence to basic principles in antitrust, which should be applied to standards and PAEs.
​​


Yee Wah Chin is counsel at Ingram Yuzek Gainen Carroll & Bertolotti LLP in New York.

The opinions expressed are those of the author(s) and do not necessarily reflect the views of the firm, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.


[1] U.S. Department of Justice and the Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property §2 (January 12, 2017).

[2] See, e.g., U.S. Department of Justice and the Federal Trade Commission, Antitrust Enforcement and Intellectual Property Rights: Promoting Innovation and Competition at 2 (April 2007).

[3] “No patent owner otherwise entitled to relief for infringement or contributory infringement of a patent shall be denied relief or deemed guilty of misuse or illegal extension of the patent right by reason of his having done one or more of the following:…unless, in view of the circumstances, the patent owner has market power in the relevant market for the patent or patented product on which the license or sale is conditioned.” 35 U.S.C. §271(d). See also, Windsurfing Intern. Inc. v. AMF, Inc., 782 F.2d 995 (Fed. Cir. 1986). In fact, a finding of patent misuse under U.S. law may be a greater burden than a finding of an antitrust violation. That is because if there is patent misuse, that patent may not be enforced at all until the misuse has been declared by a court to have been purged. E.g., Motion Picture Patents Co. v. Universal Film Mfg. Co., 243 U.S. 502 (1917). In effect, the patent holder has lost the patent until the misuse is purged. In contrast, if there is an antitrust violation, the patent holder may owe the injured party treble damages, and be subject to an injunction requiring it to cease the violation. 15 U.S.C. §§ 15, 26. The patent holder still may otherwise fully practice and enforce the patent.

[4] 547 U.S. 28 (2006).

[5] Standard Oil Co. of New Jersey v. United States, 221 U.S. 1 (1911).

[6] Northern Securities Co. v. U.S., 193 U.S. 197 (1904).

[7] Swift & Co. v. U.S., 196 U.S. 375 (1905).

[8] While the U.S. agencies continue vigorously to prosecute cartels, some of which are immense durable multi-national arrangements, obtaining record-breaking fines, these latter-day combinations affect smaller parts of the U.S. economy than their predecessors.

[9] See, e.g., Hou Liyang, Qualcomm: How China has Invalidated Traditional Business Models on Standard Essential Patents, Journal of European Competition Law & Practice (2016) at 4 doi: 10.1093/jeclap/lpw064 http://jeclap.oxfordjournals.org/content/early/2016/08/16/jeclap.lpw064.full; Sokol, D. Daniel and Zheng, Wentong, FRAND (And Industrial Policy) in China (May 5, 2016) at 18. Cambridge Handbook of Technical Standardization Law, Vol. 1: Antitrust and Patents (Jorge L. Contreras, ed., 2017 (New York: Cambridge Univ. Press)) Forthcoming, University of Florida Levin College of Law Research Paper No. 16-35, https://ssrn.com/abstract=2776235.

[10] 384 U.S. 563, 570-71 (1966).

[11] 525 U.S. 128 (1998).

[12] Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 456 (1993).

[13] U.S. DOJ/FTC Antitrust Guidelines for the Licensing of Intellectual Property, Jan. 12, 2017 (2017 IP Guidelines).

[14] Federal Trade Commission, Patent Assertion Entity Activity: An FTC Study, October 2016.

[15] Press Release, DOJ and FTC Seek Views on Proposed Update of the Antitrust Guidelines for Licensing of Intellectual Property, August 12, 2016.

[16] Discussed further, infra.

[17] Estaban Burrone, Standard, Intellectual Property Rights (IPR) and Standards-setting Process, http://www.wipo.int/sme/en/documents/ip_standards_fulltext.html#P4_83

[18] See, e.g., Allied Tube and Conduit Corp. v. Indian Head, Inc., 486 U.S. 492 (1988), where the Supreme Court affirmed the §1 liability of a member of a fire safety association for influencing the association to adopt a biased safety code to benefit its own product and disfavor competing products.

[19] In re Dell Computer Corp., 121 F.T.C. 616 (1996); In re Rambus, Inc., No. 9302 (FTC 2009), http://www.ftc.gov/os/adjpro/d9302/090512orderdismisscomplaint.pdf; Motorola Mobility LLC and Google Inc., File No. 121 0120, Docket No. C-4410 (July 23, 2013) (Order and Decision); Robert Bosch GmbH, FTC File Number 121-0081, Docket No. C-4377 (Apr. 23, 2013) (Decision and Order).

[20] Id.

[21] See U.S. Dep’t of Justice and U.S. Patent and Trademark Office, Policy Statement on Remedies for Standard–Essential Patents Subject to Voluntary F/RAND Commitments, 6 (Jan. 8, 2013); discussion at DOJ/FTC December 2012 workshop on PAE activity, http://www.ftc.gov/news-events/events-calendar/2012/12/patent-assertion-entity-activities-workshop.

[22] See Microsoft Corp. v. Motorola, Inc., 795 F.3d 1024 (9th Cir. 2015) (FRAND contract enforced, enforcement of German injunction enjoined, and FRAND royalties determined); In re Innovatio IP Ventures, LLC Patent Litig., 921 F. Supp. 2d 903 (N.D. Ill. 2013); Realtek Semiconductor Corp. v. LSI Corp., 946 F. Supp.2d 998 (N.D. Cal. 2013) (enforcing FRAND contract and granting preliminary injunction against enforcement of ITC exclusion order).

[23] Eastern Railroad Presidents Conference v. Noerr Motor Freight, 365 U.S. 127 (1961); United Mine Workers v. Pennington, 381 U.S. 657 (1965).

[24] 508 U.S. 49 (1993). 

Friday, September 22, 2017

Work description


  • A fair approximation of the major areas of my practice and their relative percentages are: Patent Litigation 40%; Patent Prosecution 40%; Others (monetization, due diligence, business development) 20%.
  • I believe two facets of my legal experience may reflect positively on this application.  
    • First, I have handled a wide array of legal matters from patent prosecution to monetization.  I think [this position] is required to supervise and participate in a wide array of legal matters.
    • Second, from at least [year] to recently, I have been ....  [This position] manages what, in essence, is [the same].  
  • I believe that my legal background over a wide array of cases and [the second thing] are factors that should be considered in connection with this application.