MIDTERM 2 REPORT

HIGH SPEED NETWORK DESIGN

Prepared by,
CHINTAN MANISH GANDHI
RED ID -818488938

Question 1:
Compute the raw physical data rate for both uplink and downlink channels at each bandwidth
range (at 2.5, 5, 10 and 20 Mhz)?
Answer 1:
As per the given data we have frame size of 16ms, which is divided into 20 slots of 800microseconds, that
gives 25 symbols per slot.
It is given that we have to use 60% of the frame time as Downlink channel and 40% of it will be the
Uplink Channel. Thus we have total 8 slots for Uplink Channel and 12 slots for Downlink Channel.
Now we have subcarrier spacing of 39.06 KHz, which gives different number of subcarriers in different
bandwidth range. Thus considering 12 Guard Band subcarriers for each bandwidth and 4 pilot subcarriers
at,
2.5MHz we have 64 subcarriers - 12 Guard Band subcarriers = 52 Data subcarriers.
Similarly for 5MHz, 10MHz and 20MHz we have 116, 240 and 500 Data Subcarriers respectively.
Thus to calculate Data rate for every bandwidth we need to calculate for every modulation scheme and
Code rate.
The formula for calculating,
DL data rate (bits/sec) per Frame = (No. of Subcarriers * Code Rate* Modulation Scheme* No. of
bits per symbol * No. of symbols per slot * No. of DL Slots)/Frame Duration.
The formula for calculating,
UL data rate (bits/sec) per Frame = (No. of Subcarriers * Code Rate* Modulation Scheme* No. of
bits per symbol * No. of symbols per slot * No. of UL Slots)/Frame Duration.
Please refer to the excel sheet for all the calculations.
Question 2:
Assume OFDMA will be used in the downlink channel and SC-FDMA will be used on the uplink
channel. Also assume that channel allocation unit in both uplink and downlink channel is defined
by a PRB (physical resource block). A physical resource block is defined as 4 subcarriers over a
period of one slot. Compute both uplink and downlink bandwidth in terms of physical resource
blocks. Provide your computations in a table listing the number of physical resource blocks per
bandwidth category in terms of PRB.
Answer 2:
In this all the channel Allocation is done through PRB (Physical Resource Block) we need to specify how
much PRBs we will be having for both Uplink and Downlink Channel.
It is given that a PRB consists of 4 subcarriers over a period of one slot. And we have different numbers
of subcarriers for different bandwidth per frame of 16ms. Also considering all the subcarriers for a given
bandwidth we need to calculate total numbers of PRB per slot and then for the entire Frame size for both
that is Uplink & Downlink Channels.
So for computing number of PRB for DL channel per Frame is,
No. of PRB per Frame = No. of PRB per Slot* Total No. of slot per frame* % of BW allotted to DL.
Also for computing number of PRB for UL channel per Frame is,
No. of PRB per Frame = No. of PRB per Slot* Total No. of slot per frame* % of BW allotted to UL.
Please refer to the excel sheet for all the calculations.
Question 3: Assume that 4 resource elements are used to transmit reference signals which are used
for synchronization and channel estimation per physical resource block. Compute the amount of
bandwidth consumed by reference signals in both uplink and downlink channels.
2

Answer 3:
We are given that there are 4 resource elements, which are used to transmit the reference signals used for
synchronization and channel estimation on per resource block.
Thus we have to calculate this value for every bandwidth.
Starting with,
No. of resource elements in one PRB = No. of symbols in one PRB * No. of subcarriers in one PRB
Now that we have total number of resource elements in one PRB. We know from that number 4 resource
elements are used to transmit reference signals in each PRB.
Now we have to calculate total number of PRBs for both UL and DL.
Total No. of PRBs in UL = Total No. of PRB per slot * Total no. of slots per frame * 0.4(40% of UL
traffic)
Total No. of PRBs in UL = Total No. of PRB per slot * Total no. of slots per frame * 0.6(60% of DL
traffic)
Now,
No. of Resource elements in UL per frame = No. of PRBs in UL * No. of Resource elements per
PRB
No. of Resource elements in DL per frame = No. of PRBs in DL * No. of Resource elements per
PRB
Also we need to calculate No. of Reference signals in UL and DL for the entire frame, which is
No. of Reference signals in UL = No. of Reference signals per PRB * No. of PRBs in UL
No. of Reference signals in DL = No. of Reference signals per PRB * No. of PRBs in DL
Now after calculating all the above values we can calculate much these reference signals would occupy
the bandwidth in both UL and DL channels. This can be computed as follows,
Bandwidth consumed by reference signals in UL = No. of Resource elements in UL * No of
Reference signals in UL channel.
Bandwidth consumed by reference signals in DL = No. of Resource elements in DL * No of
Reference signals in DL channel.
Please refer to the excel sheet for all the calculations regarding the bandwidth occupied by the reference
signals in both UL and DL for every bandwidth.
Question 4: Assume that 2x2 MIMO is employed in downlink and 1x2 MIMO is employed in uplink
communication. Compute the raw data rate for both uplink and downlink channels.
Answer 4:
It is asked that MIMO technology is implemented in both Uplink and Downlink communication.
It is given that 2x2 MIMO is used in Downlink communication. Thus there exists 2 transmitting antenna
at eNB and two receiving antenna at UE to download the data.
Thus to Raw data rate for DL channel the Raw data calculated in question 1 should get doubled.
That is, 2x2 MIMO DL Raw data rate = 1x1 DL Raw data rate * 2
Now 1x2 MIMO is employed in Uplink communication, thus there exists only one transmitting antenna at
UE and two Receiving antennas at receiver side i.e. at eNB.
3

So the Raw data rate calculated in question 1 for uplink will remain same even if we employed 1x2
MIMO for Uplink communication.
Please refer to the excel sheet for all the calculations regarding the employment of MIMO in both UL and
DL for every bandwidth and different modulation scheme and coding rate.
Question 5: Discuss the design issues related to uplink channel. Discuss how bandwidth will be
allocated to users and communicated to the users. Discuss how bandwidth requests will be
communicated by the user equipment to the base station.
Answer 5:
The basic design issues related to the uplink channel are PRB allocation and Synchronization.
As the design is coordinated channel type, each UE needs to have some kind of synchronization with
eNB, if the UEs are not synchronized with eNB, UE wont be able to detect the frame start and would not
be able to properly set the scheduling of the data during the uplink channel in the allotted slot instead it
will send the data in different slot resulting into collision.
Another problem is PRB allocation indexing, which brings the issue that how UE will be able to detect
the start of the PRB or where exactly the allocated PRB resides in the particular slot or the frame.
Bandwidth allocation to user:There are number of ways with which bandwidth can be allocated and communicated to users.
The best method to use is to have a coordinated channel where each user will get a particular time slot to
transmit their data.
Thus in contrast of the bandwidth allocation in LTE, in our design a MAP (Medium Access Planner)
signal is sent in the downlink channel by eNB (Base station). This MAP will have a frame structure that
will have fields defining which user will send when and how many PRBs are allocated to particular user
(UE) along with the starting point of the PRB for transmission, i.e. from what PRB # the user will start to
transmit the data. The eNB will also specify what modulation scheme and code rate should use to transmit
data on uplink channel in order to have successful transmission. The MAP packet is updated when the
requirements of the UE for the required number of PRB increases or decreases as per need.
The UE will send the Reservation signal asking amount of PRBs for the transmission of uplink traffic
and also specifies the priority of data type, in response to this the updated MAP is sent again by the eNB
in the downlink channel to communicate to UE that the request is accepted and the resources are allocated
to transmit the data in the successive frame.
The issue is that, there will be many users who will try to get the access of the channel, the MAP will
assign the PRBs according to the priority of the data transmission of each UE, this way the UEs with
parameterized data will get more preference over the UE having non-parameterized (Asynchronous) data
to get access of uplink channel.
The MAP packet will be send during the initial slot of the downlink channel, and whether this packet
should be sent in every frame or not will be the decision of eNB depending upon the traffic, but in our
design MAP will be transmitted whenever there is update to the network, i.e. a new user has joined or left
the network or if the UE requires more PRBs or less.
Bandwidth request by UE:The bandwidth request will be made by UE in the uplink channel to access the channel with the help of
the signal known as Reservation signal in Physical Uplink Control Channel (PUCCH) or using
piggybacking with data on Physical Uplink Shared Channel (PUSCH). This frame will have the
4

information regarding the amount of PRBs needed by the user to transmit their data and also the priority
of the data.
The UE will listen during the downlink on the Physical Downlink Control Channel (PDCCH) for the
MAP signal. Once the MAP is arrived, the UE will extract the MAP signal to obtain the information
about its transmission time and modulation scheme to be used to send data in uplink channel.
Every UE has predefined allotted PRBs by eNB, there exists a bias value which will define number of
allocated PRBs every UE initially. Thus if UE needs less number of PRBs it will just inform eNB about a
value to be subtracted from the bias value, and will grant the remaining # of PRBs to UE, if UE requires
more PRBs than the allotted bias value of PRBs, then the value is added to the bias value thus more PRBs
are given to that UE. This way the UE can ask for more or less number of PRB according to the need.
Reservation signal also consist of the information field which specifies the channel quality indication
sent from UE to eNB to inform what modulation scheme and coding rate UE can support. UE also sends
the information of ACK/NACK to eNB using HARQ protocol.
The eNB will receive many Reservation signals from different UE, but it can accommodate only few
requests in one frame, i.e. only few UEs can transmit their uplink traffic but those request which are not
granted to have the access of the channel by eNB, they will wait for next uplink duration to send their
Reservation signal to get access of channel from eNB.

Question 6:
Discuss how downlink channel will be allocated to individual users. How will channel allocation be
communicated to the users?
Answer 6:
In our design, if UEs wants to communicate within themselves they have to send all the data to eNB
during the Uplink Channel, and all that data for the destination UEs is then transmitted by eNB in
downlink channel. Thus eNB is aware all the time that how much data is going to be transmitted by UEs
in the Uplink channel, and how much information it has to transmit to the UEs during the Downlink
channel. Thus the scheduler in eNB does the bandwidth allocation to appropriate user but also which
radio bearer to serve to that particular UE. Thus eNB will just sent all the data which is received during
the UL of UEs on the DL channel. Sometimes there are important data to be served before the
5

asynchronous type of data. Thus those data will be sent prior to any other data to the particular UE who
has requested that data.
Channel allocation to the UE will be communicated through the MAP (Medium Access Planner) signal in
any slot in the downlink channel, is purely the decision of eNB. As MAP signal occupies only some of
the PRBs in Downlink, that will depend upon number of UE Information Entity (UIE). It is not
transmitted in every slot and also not in every frame, but it is sent only when it is required i.e. when there
are some updates in the network, which overall leads to less overhead. The MAP will have a generic
structure which will specify which UE will transmit the data, with what modulation scheme and coding
rate and also the number of PRBs allocated to particular UE for transmission of Uplink Traffic in uplink
channel.
Question 7:
Suggest a generic and efficient frame structure for communicating channel allocation and usage
information to user equipment. Discuss the frequency of these frames along with the impact on the
bandwidth and latency observed in the channel.
Answer 7:

Frame = 16ms

DL Downlink Slot
UL Uplink Slot
GI Guard Interval slot, transition from downlink to uplink which is 16microseconds long which is
part of the Downlink slot. (In above figure it is shown GI takes the whole slot but it is just part of
last DL slot)
Information fields:
1. Number of UEs - This field will specify how many UEs will transmit their uplink traffic in
coming frame. The length of MAP signal will depend on the number of UEs.
2. Packet Type This field will specify the code sequence through which UE will come to that the
incoming signal is MAP signal.
3. UE ID This field will specify the 8 bits User Identification ID.
4. Number of allocated PRBs This field will specify the total no. of PRBs are allotted to UE.
6

5. Slot Number These 8 bits will define the slot number in which UE will transmit the uplink
data.
Thus there are total 20 slots in the frame, so eNB will assign 8 bit number to each slot. So now if
eNB wants to say UE should send its uplink traffic particular uplink slot, it will send the number
sequence representing that slot in downlink.
6. Start PRB Number This sequence will define from what PRB the UE will start transmit the
data.
There are total X no. of PRBs in the frame, thus to specify which PRB to be used, indexing is
done through this field, where each PRB in frame is represented by this 8 bit number. So if eNB
wants to allot particular PRBs in particular slot, eNB will send the number representing start of
PRB and from that PRB onwards amount of PRB will be occupying by the UE is specified in 4th
field of the MAP signal.
7. MCS This field will specify what Modulation scheme and coding rate UE should use to
transmit their uplink traffic.
This MAP signal can be transmitted in any downlink slot; this decision is of eNB. In my design the MAP
signal is only transmitted when there is an update to the network. The number of PRB occupied by this
signal will depend upon how many UEs wants to send their data. Thus if there are more UEs the MAP
signal will require more number of bits which ultimately will increase the overhead. And if all the users
are not allotted the access the uplink channel in the single frame, then another MAP signal has to be sent
in another frame giving other UEs to transmit their data, this will result in latency to send the data.
Question 8:
Suggest a frame structure for communication reservation requests, Ack and NACK information
from user equipment to the base station (Central Coordinator). Discuss how this frame will be
mapped to the uplink channel PRBs.
Answer 8:
In our design each UE will communicate with eNB using Reservation signal. The below figure shows
the structure of Reservation signal,

1. Packet Type This field will have a code sequence of 4 bits through which eNB will be notified
that the incoming signal is Reservation signal.
2. UE ID This 8 bit number will have the UEs Identification number, so that eNB will come to
know that which UE has sent the Reservation signal.
3. CQI (Channel Quality Indicator) This field will specify what Modulation scheme and coding
rate UE can support to transmit Uplink traffic.

4. # of PRB + or : This is 8 bit number which will specify how many more or less PRBs are
required by UE from the initial bias value, i.e. Bias Value = Bias Value + # of more PRB or
Bias Value = Bias Value - # of less PRB . The eNB in that case will add or subtract that value
from the Bias value and will grant that many PRBs to that particular UE.
Say for example the initial value of bias is 20. Now if UE wants 15 more PRB, this field will
have number which will be
Bias Value = Bias Value + # of more PRB
35 = 20+15
Thus field will include the number specifying the value 35.
Similarly if UE wants to decrease the allotted number of PRBs from 20 to 10, it will send,
Bias Value = Bias Value # of less PRB
10 = 20 10
Thus field will include the number specifying the value 10.
The information sent in this field will be compared with the bias value at eNB, thus if the
compared value is positive at eNB, it will grant that many PRBs which the compared value has,
in contrast if the compared value is negative than, then eNB will cut down the amount of PRBs
given to that UE by the amount of compared value.
5. ACK/NACK This information is sent from UE to eNB in response to the received downlink
data.
6. Priority This information will define the priority of the data UE want to transmit. UE with
higher priority data will be given more preference over the other UEs data having low priority.
This Reservation signal can be transmitted in two ways. It can be send either on PUCCH
(Physical Uplink Control Channel) or it can be send by piggybacking with the data on PUSCH
(Physical Uplink Shared Channel). In my design it is transmitted on PUSCH piggybacked with
the data by every UE in the very first PRB allocated to the UE during the initial symbols in PRB.
The below figure shows how Reservation signal will be mapped on the Uplink channel,

The above figure shows that how Reservation signal is sent on uplink channel piggybacking with the data.
For simplicity Reservation signal of only 1 UE has been shown in the figure. The Reservation signal will
occupy certain symbols in first PRB allocated to that UE, the rest will be used for data. If there are
multiple UEs occupying the same slot, there can be more than one Reservation signal in one slot.

Question 9: Based on your answer to questions 5-8, compute the MAC overhead related to channel
allocation plans communicated to user equipments and bandwidth allocation requests
communicated to the base station.
Answer 9:
There exists three types of MAC overheads in the network and they are,
1. MAP signal
2. Reservation signal
3. Reference signal in Reference Element
The overhead for each is calculated as follows,
Let us assume the bandwidth used is 2.5MHz, Modulation scheme to be 64QAM and coding rate of ,
and number of users in network to be 30.
No. of bits per symbol in per subcarrier = No. of bits per symbol for 16QAM * Code rate
= 4 bits/symbol * 3/4 = 3 bits/symbol per subcarrier
= 3 bits/Resource element
No. of Reference signals per PRB = 4R.S/PRB
No. of bits for Reference signal per PRB = 4*3 = 12 bits/PRB
No. of bits of Resource elements per PRB = 100 Resource elements (R.E) * 3bits/ R.E.
= 100*3 = 300 bits/PRB
No. of PRBs in UL Channel = 104
No. of PRBs in DL Channel = 156
No. of bits of Reference signals in UL Channel = No. of PRB in UL * No. of bits of Reference
signals/PRB
= 104 * 12 bits/PRB
= 1248 bits
No. of bits of Reference signals in DL Channel = No. of PRB in DL * No. of bits of Reference
signals/PRB
= 156 * 12 bits/PRB
= 1872 bits
No. of bits for MAP (for 30 users) = No. of bits per MAP signal * No. of UEs + 8 bits (specifying for
# of UEs)
= 40 * 30 + 8 bits
= 1200 +8 = 1208 bits.
No. of bits for Reservation signal (for 30 users) = No. of bits per Reservation signal * No. of UEs
= 32 * 30
= 960 bits
Total MAC overhead = Total No. of bits for Reference signal (UL or DL) + Total No. of bits for
MAP signal or + No. of bits for Reservation signal.
Total MAC overhead for DL = Total No. of bits for Reference signal + Total No. of bits for MAP
signal
= 1872 + 1208 bits = 3.080 Kbits
Total MAC overhead for UL = Total No. of bits for Reference signal + Total No. of bits for
Reservation signal
= 1248 + 960 bits = 2.208 Kbits

Question 10: Discuss the life cycle of one asynchronous data packet of length 1500 bytes in the
uplink and downlink channel. Suggest a generic frame structure to for both uplink and down link
channels.
Answer 10:
Let us now discuss the life cycle of the packet for both Uplink and Downlink channel.
Life cycle of the packet in DL:-

Downlink
Flow

In Downlink channel, the packet is received from the network to the PHY layer of the UE. Thus UE can
deliver single packet or multiple packets received on PHY layer to MAC layer, depending on the
scheduling and modulation.
There are functions which each layer performs,
1. MAC Layer - It is responsible for managing the hybrid ARQ function, which is a transport-block
level automatic retry. It also performs the transport as a logical mapping. Types of logical
channels include control channels (for control plane data) and traffic channels (for user plane
data). Transport channels are in the transport blocks at the bottom of the MAC. They represent
data transfer services offered by the PHY and are defined by how the information is carried,
different physical layer modulations and the way they are encoded.
2. RLC Layer - It performs segmentation and reassembly and operates in three modes: transparent
mode (TM), acknowledged mode (AM) and unacknowledged mode (UM). These are used by
different radio bearers for different purposes. The RLC provides in-sequence delivery and
duplicate detection. RLC Segmentation is the segmentation process involves unpacking an RLC
PDU into RLC SDUs, or portions of SDUs.

10

3. PDCP Layer: PDCP functions in the user plane include decryption, ROHC header
decompression, sequence numbering and duplicate removal. PDCP functions in the control plane
include decryption, integrity protection, sequence numbering and duplicate removal. There is one
PDCP instance per radio bearer. The radio bearer is similar to a logical channel for user control
data.
PDCP Header Compression is important because VoIP is a critical application for LTE.
Ciphering and integrity are two most important functions in PDCP. They are used for security
purpose that UE which is not part of the network can listen the Data.
Lifecycle of packet in Uplink Channel:

Uplink
Flow

1. PDCP Layer: PDCP functions are symmetrical for the uplink and the downlink. The functions
are the same for the header compression and encryption, but they occur in reverse order. Uplink
processing includes header compression and encryption.
2. RLC Layer: The uplink process concatenates rather than segments the SDUs into transport
blocks. Segmentation is only done when it is needed to fit SDUs into a transport block.
Concatenation is the process of packing an RLC SDU into a size appropriate for transport blocks.
3. MAC Layer: Uplink functions include random access channel, scheduling, building headers and
transport format selection.

11

Generic Frame Structure:

Uplink
Flow

Question 11:
Assume that one PRB is allocated to a User per frame time. Compute the latency of a packet of
length 1500 byte in the uplink and downlink channel.
Answer 11:
Here it is given that one PRB is allocated per user per frame.
Assume that bandwidth is 2.5MHz, modulation scheme as 16QAM and coding rate of 3/4
We know No. of bits per PRB = No. of bits per symbol *
= 4 bits/symbol * = 3 bits/symbol per subcarrier
Thus No. of bits per PRB = 100 * 3bits/symbol per subcarrier
= 300 bits/PRB
Now for 1500 bytes of data = 1500 * 8 bits = 12000 bits of data.
Therefore No. of bits transmitted per frame = 300 bits.
Thus total no. of frames required to send 1500 bytes of data = No. of bits of data / No. of bits per frame
= 12000 / 300 = 40 frames.
As we know each frame is 16ms long, and we require 40 frames to transmit 1500 bytes of data.
The latency of a packet of length of 1500 bytes = 40 * 16ms/frame = 640ms
Question 12:
Provide a general discussion on integration of the access network with the local area network you
designed earlier. What all are the design and implementation issues regarding this integration.
Answer 12:
The earlier design was WhyNet Network (a type of local area network), which was designed to be a
coordinated network (Packet based network) where each WhyNet node will have certain time slots in the
cycle time to transmit their data to other WhyNet node. The access to use the resources in the network is
granted by the WhyNet Channel Coordinator.
12

This Channel coordinator has particular job to do in the network. The function of Channel coordinator is
to allocate a slot to any WhyNet node that is part of the network, to transmit its data. In WhyNet, the
communication is not through the Channel coordinator but it is between nodes to nodes. Thus the primary
role of Channel coordinator is to provide channel resources (in terms of slots) to WhyNet nodes to
transmit their data. The Channel coordinator does the allocation of slots of cycle time using MAP Packet
that will specify which user will send their data in which time slot.
There are two types of data transmitted in the network, Isochronous (media files, VoIP, etc.) and
Asynchronous data. The bandwidth allocated for Isochronous data is 80% and for Asynchronous it is
20%.
Design and implementation issues,
1. Synchronization of Channel Coordinator and eNB (Hardware or software synchronization).
2. Problem related to asynchronous and isochronous type of data, there is nothing like this in the
access network, there exist only one type of data which has certain priority.
3. The Channel coordinator has to send data to eNB at certain frequency, modulation scheme and
coding rate so that eNB can able to understand the packet.
4. Channel coordinator and eNB both should posses some kind of buffer system, so that in case of
network failure on either side the incoming data can be stored, this will make both the Channel
coordinator and eNB to have more complex protocols.
Here are some methods through which we can overcome some of the issues stated above,

Integration of local area network and Access network

To integrate Local area network (WhyNet network) and Access network,

Also the Channel Manager at WhyNet side will request the Resource allocation with the RAS (Resource
allocation signal) to eNB to transmit the data, if node in LAN wants to communicate with the UE in
Access network. In response to this the RAS signal if eNB has granted the request of Channel
coordinator, it will send RAR (Resource allocation Response) signal to Channel coordinator.
Another role which the Channel Manager will perform is to send Pseudo ACK in case of the network
failure on either of the side, the Channel manager will send to corresponding Channel manager to notify
that the network is bad but it has received the data and you can send more data but the destination node
will get the data once the channel is clear, thus all the data received on Channel manager will be stored in
the buffer. This mechanism will ensure that there is a continuous flow of data and bandwidth is not lost
and does not result into latency.

14

References:
1. White Paper on LTE protocol overview
http://www.freescale.com/files/wireless_comm/doc/white_paper/LTEPTCLOVWWP.pdf
2. IEEE paper on The LTE-Link Layer Design by Anna Larmo, Mangus Lindstrom.
3. IEEE paper on Spectral Efficiency assessment and Radio Performance between LTE and Wimax
by Carsten Ball and Thomas HIndelang.

15