OpenAirInterface.org Specifications

Scope

The present document specifies the architecture of the following components of a OpenAirInterface network

1. Radio Network Topology and Components
2. Physical Layer Procedures - Mesh Topology
3. Physical Layer Procedures - Cellular Topology
4. Layer 2 Protocols

Radio Network Topology and Components

This section describes the different components which constitute the OpenAirInterface. Two types of network topologies are supported, mesh and cellular. The mesh topology is depicted as:

OpenAirInterface Mesh Topology

The cellular topology is depicted as:

OpenAirInterface Cellular Topology

The mesh topology contains Clusterheads (CH) and Mesh Routers (MR) whereas the cellular topology contains Clusterheads/NodeB (CH/NodeB) and User Equipment (UE). The main difference at the physical layer between the two topologies is that direct communications between UE is not permitted in a cellular topology. Other major differences exist at layers 2 and 3. In both topologies, MR/UE can be connected to more than one CH at any time, if radio connectivity is possible.

The two types of physical networking devices are specified as follows:

• Cluster-head (CH or NodeB): A cluster-head can be defined as a node with a maximum visibility/connectivity in terms of number of nodes in its neighborhood. From the point-of-view of the MAC layer, it assumes the fine-grain management of radio resources in the cluster (cell). It determines the frame partitioning (see TTI and Modulation Parameters) and bandwidth allocation and communicates this information to nodes in the cluster through a beacon. At the physical layer, the cluster-head provides mechanisms for timing and carrier frequency synchronization. The primary role of the CH is to manage radio resources in their cluster. The cluster is defined as the set of nodes which are characterized by one-hop connectivity with the clusterhead. One-hop connectivity is further defined as the capacity to reliably receive and transmit basic signaling channels with the clusterhead using at least at the lowest datarate communication mode. Reliable communication is defined by a transmission which falls below a maximal error probability threshold. CH can only be connected to MR on the same frequency-carrier since they use the same temporal resources as other CH. Thus direct CH<->CH communication is not possible on the same frequency carrier. The downlink (CH -> MR) signaling channels allow for the CH to schedule transmission of labels (in the form of time and frequency mappings on the radio resource) which each carry different types of traffic throughout the mesh network. The Uplink (UL) signaling channels (MR -> CH) are used for relaying bandwidth requirement indicators and channel quality measurements from nodes within the cluster. These feed the scheduling algorithms residing in the CH and allow for proper resource allocation satisfying quality-of-service (QoS) negotiations carried out using Layer 3 (L3) signaling.

The CH further provides mechanisms for measurement reporting to L3 (for routing, QoS management, labeling, etc...). This is achieved by a set of signaling channels which relay measurement information (UL) for the nodes in the cluster to the CH. The CH processes these raw measurements into a form which is expected by L3 mechanisms. Some CH can assume the role of network synchronization by sending special synchronization pilots (see Section 1.1.5). These will be called Primary CH when network synchronization is achieved using this method. Other CH using this method are called Secondary CH.

• Regular node/Mesh Router (UE): All nodes have the ability to play at the same time the role of a host and of a router, although this functionality is not activated in a cellular topology.

The primary role of an MR is to interpret the scheduling information from the CH on the DL signaling channels in order to route the traffic corresponding to the scheduled labels on the allocated physical resources. MR can be connected to other MR (direct link) in the same cluster. MR can also be connected to more than one cluster at the same time. It is also expected to using the UL signaling channels to relay measurements to the CH with which is connected. A secondary role of some MR is to search, on behalf of the CH, for isolated nodes which need to be connected to the mesh. These MR use a special signaling resource (MRBCH) to exchange basic topological parameters with the IN which then results in overall network topology updates. If several IN are contending for access with the cluster, joint processing of the requests will be performed by the mesh during topology adjustments. The most likely nodes to assume this role will be those at the extremities of the mesh.

Either type of equipment can also assume the role of relay/gateway to a secondary network. `

In mesh topologies some nodes (CH or MR) assume the role of and edge router (from a layer 2/3 perspective). An edge router is either a CH or MR with an IP interface to another network. The role of ER is to aggregate traffic (ingress) from IP flows to MPLS-like labels for transmission in the mesh. On reception it must demultiplex traffic (degress) from MPLS-like labels to IP for traffic exiting the mesh. Edge routers, potentially all CH and MR, must have MAC-layer interfaces to IP in order to perform these functions.

Both OpenAirInterface topologies require Network Synchronization (NS) at least between adjacent clusters. This must be on the order of a few microseconds. Three mechanisms are supported to ensure NS. Firstly, a secondary synchronization source (e.g. GPS) can be used as a common time reference by all nodes. Secondly, one CH (Primary CH) in the network use a special synchronization signal which has longer range than the range of communication, in order to cover the region which a common time reference. This is suitable for small networks. Finally the method of distributed relaying of synchronization is possible. This is a method by which all nodes propagate a time reference. Nodes switch between reception (for timing acquisition and tracking) and transmission of the reference. This guarantees coverage of network synchronization over long distances in the absence of a secondary synchronization source.

Layer 2 Protocols

Layer 2 is structured as below. It comprises:

• A IP/MPLS networking device (NAS DRIVER) responsible for provision of IP/MPLS layer services to Layer 2 and vice-versa
• An MPLS label-switching entity (NAS MPLS) responsible for routing/forwarding within the mesh network (MESH topology only)
• A Radio resource control (RRC) entity responsible for MAC layer signalling for configuration of logical flows (labels) and retrieval of measurement information.
• A Radio Link Control (RLC) entity which is responsible for automatic repeat request protocols (ARQ) and IP/MPLS packet segmentation
• A convergence protocol (PDCP) responsible for IP interface and related functions (header compression, ciphering, etc.)
• A scheduling and multiplexing unit (MAC Layer (MAC) Specifications) responsible for the mapping between logical channels (labels and control-plane signalling) and transport channels. It implements the interface with the PHY, which is the collection of transport channels as well as a primitives for collection of PHY measurements and configuration of PHY parameters.

Global View of OpenAirInterface Protocol Stack and Communication Primitives

These entities are described in the following subsections.

Radio Resource Control (RRC)

The radio resource control entity is responsible for the L2 signalling implementing the radio channels establishment. It also implements the control of measurement procedures described in Section 1.3.4. Its internal state machine controls the basic procedures for startup, monitoring of synchronization through the measurement system and update of the nodes role in the network (Sections 1.3-0).

RRC is responsible for configuration of all MAC entities (and PHY via MAC), both dynamic (during label establishment) and static (control channels). This functionality is in response to event occurring in the interaction with L3 and based on dynamic measurements of radio quality.

RRC signalling makes use of DCCH, CCCH and BCCH for transport of the various protocols.

Radio Link Control (RLC)

RLC segments IP packets. The segment size is configurable for each QoS class and is signalled by higher layers during route establishment. The sizes are chosen based on the granularity of the underlying MAC/PHY resources (transport blocks).

RLC is responsible for ARQ and indexing of SDUs from the user traffic and signalling SDUs from RRC. The SDU inputs from LS form the set of radio bearers, and those from RRC the set of signalling radio bearers. It has two modes of functionality: acknowledged and unacknowledged. Each logical channel can have an associated ARQ process which is managed by RLC. The ARQ mechanisms are based on Release 6 3GPP RLC (25.8xx). The interface with RRC for configuration is not yet described. The interface with MAC is designed such that data for each logical channel is buffered in data queues, whose occupancy can be measured by the MAC scheduling entity.

MAC scheduling Entity (MAC)

The MAC entity is responsible for scheduling control plane and user-plane traffic on the physical OFDMA resources. On transmission, the inputs to this entity are connected to data queues originating in the RLC layer which form the set of logical channels. The control plane traffic is represented by logical channels which form the interface with the RLC. Logical channels contain both user-plane (originating in the IP/MPLS entity via the PDCP entity) and control-plane traffic (originating in the RRC entity). MAC layer specifications are found in MAC Layer (MAC) Specifications. The MAC is responsible the transport channel interface which exchange data (MAC SDUs) and PHY measurement data for RRC measurement procedures.

Physical Layer Procedures - Mesh Topology

Physical Layer Procedures - Cellular Topology

Power-on procedures of a CH

This clause briefly describes the power-on procedure of a CH. The CH RRC receives basic cell configuration information (CHBCH/RACH configuration) from L3 Radio Resource Management (RRM) and configures the MAC and RLC layers. The MAC, in turn, configures the static paramters of the PHY. Upon completion of this initialization phase it enters the steady-state mode.

Power-on procedures of a UE

This clause briefly describes the power-on procedures of a UE which is managed by the UE-RRC state machine. The first function is to search for at least one existing CH in range which is under the responsibility of the PHY. This procedure attempts to analyze the received signal power over a pre-defined time period. If a signal is detected on the desired carrier, the node attempts PHY synchronization using the pilots of the candidate CHs. The postulated frame start position is used to demodulate the CHBCH resources. If the PHY returns an error-free CHBCH with acceptable receive quality, the node is said to be pre-synchronized to the CH. It then attempts to decode the CHBCH of the rest of the CH in range. At the end of this stage, it has a list of acceptable CH. This procedure is repeated periodically, until active communication is sought with the network.

Upon passing to the state of active communication, it demodulates the CHBCH continuously from the CHs to obtain the UL-CCCH configurations as well as MCCH/MTCH. Once configured, it attempts to establish a connection (connection request) with each of the candidates using the RACH resources of each CH. Upon successful completion of the association procedure it is said to be synchronized to the CH and enters the steady-state mode for each CH.

CH Steady-state operation

In TTI N, the CH transmits all DL flows as determined by the CHBCH scheduler during the end of TTI N-2. Furthermore, it detects the UL-SACCH for all UE flows and the UE-SACH for all flows. It also detects the RACH (CCCH). At the end of each TTI, once it has received all feedback indicators (channel and queuing), it invokes the CHBCH scheduling entity to determine the allocations for TTI N+2. The sequence of operations after receiving flows from TTI N (towards the end of TTI N+1) is:

Entity: RRC TX

1. generate BCCH and CCCH to be conveyed in TTI N+2
2. program new logical channels and measurement procedures starting in TTI N+2 based on L3 signaling requests
3. generate signaling radio bearers (measurement requests, UE radio bearer/logical channel configuration) for TTI N+2, and invoke RLC

Entity: MAC TX

1. Invoke MAC Scheduler for TTI N+2 allocations : compute DL_SACCH_PDU and UL_ALLOC_PDU for N+2, corresponding to PHY allocations in N+3.
2. Generate CHBCH PDU for TTI N+2
3. Retrieve RLC SDUs and generate DL_SACH for TTI N+2
4. Generate MACPHY_DATA_REQ for TX transport channels in TTI N+2

MAC RX

1. Generate MACPHY_DATA_REQ for RX transport channels in TTI N+2 (based on previously scheduled UL_ALLOC_PDU received by UE in TTI N, decoded by end of TTI N+1)
2. Invoked by PHY through macphy_data_confirm, the received flows are routed to RLC data queues and MAC signaling information is stored for MAC TX scheduling in next TTI. The MACPHY_DATA_IND primitive (invoked by PHY) also provides CH RX measurements information in an UL_MEAS structure.

RRC RX

1. Retrieve RACH and process association requests
2. Retrieve signaling radio bearers from RLC
3. Retrieve CH RX measurements from PHY/MAC

This sequence is invoked at the end of each TTI by the system scheduler.

UE Steady-state operation

In the steady-state of TTI N, the UE PHY demodulates the CHBCH. The CHBCH PDU is then available during TTI N+1 for the MAC. to determine the allocations of the CH and itself in TTI N+2. Based on the decoded information, its scheduling entity generates the transmission for the next TTI and configures the PHY to demodulate the data for which it is destination in the current TTI. The UE RRC acts on PHY/MAC measurements to maintain proper synchronization and received signal quality, for example by detecting a loss of connection of degradation of service. The sequence of operations at the end of TTI N is

Entity: MAC RX

1. Parse CHBCH_PDU
2. Generate macphy_data_req for RX transport channels in TTI N+2 (based on previously scheduled UL_ALLOC received by UE in TTI N, decoded by end of TTI N+1)
3. Invoked by PHY through macphy_data_confirm, the received flows are routed to RLC data queues and MAC signaling information is stored for MAC TX scheduling in next TTI. The reported PHY RF measurements (i.e. in DL_MEAS structure) are processed and used to generate UL_SACCH_FB.
4. Process measurements for RRC measurement reports and invoke mac_meas_ind for each logical channel requiring a measurement report.

Entity: RRC RX

1. Retrieve BCCH and CCCH and generate association requests
2. Retrieve signaling radio bearers from RLC
3. Retrieve UE RX measurements from PHY/MAC for L3 measurement reporting

Entity: RRC TX

1. program new logical channels and measurement procedures starting in TTI N+2 based on L3 signaling requests signaled by CH-RRC
2. generate signaling radio bearers (measurement reports, configuration ACKs) for TTI N+2, and invoke RLC

Entity: MAC TX

1. Invoke MAC multiplexer for TTI N+2 allocations : compute UL_SACCH_FB and UL_SACCH_PDU for N+2, corresponding to PHY allocations in N+3.
2. Generate CHBCH PDU for TTI N+2
3. Retrieve RLC SDUs and generate UL_SACH for TTI N+2
4. Generate MACPHY_DATA_REQ for UL_SACH in TTI N+2

QoS Measurement Procedures

CH RRC manages L3 measurement reports at L2 for nodes within the cell. Measurement reports are exchanged between UE and CH using a logical channel (DCCH) for topological control signaling, and edge routers can provide these measurements to IP Since the CH scheduler has access to low-level PHY measurements, the MAC layer is responsible for measurement reporting on behalf of the PHY and itself. The CH obtains raw measurements of all links in the cell. RRC acquires these measurements from MAC scheduling entity. Measurements are processed in nodes to the degree required for higher level services. For example nodes will extract link quality (rate/delay) indicators from low-level services (MAC to L3 measurement messages) which are transported using special signaling flows offered by the MAC. This is then used for L2.5 topology maintenance (radio-bearer (re)-assignment). Edge routers will extract L2.5 measurement information on labels to provide IP with quality indicators.

The interface with RRC for measurement reports is very similar to existing Release 6 HSPA. The types of measurements are:

• periodic (or one-shot) with configurable reporting interval and total number of measurements
• event-driven - in order to handle degradation of QoS level or loss of connection.

The available measurements for CH RRC(L2) are:

• RSSI (dBm) on physical resources corresponding to logical channel.
• Average SINR (dB) on physical resources corresponding to logical channel.
• Average number of transmission rounds (times 10) on transport channel associated with logical channel.
• Average residual block error rate (times 1000) on transport channel associated with logical channel (after HARQ!).
• Actual Spectral efficiency (bits/symbol times 10) of transport channel associated with logical channel.

---