Network Layer/Internet of Things (Introduction)

 Here is a diagram of the TCP/IP Protocol Stack. 

The Internet Protocol, or IP, is a Basic Delivery System for packets between devices that are not necessarily directly connected and it is a universal standard. 

The IP network is connectionless. In the network, each packet is treated independently from the others, simplifying routing. They are unreliable, since packets can be lost, delayed, and delivered out of order. They are also best-effort, since here the network does its best without guarantees. We can leave the network as simple as possible and push the complexity to the endpoints. Routers are as efficient as possible, and the intelligence is at the end of the devices. 

Now, comparing IP addresses and MAC addresses, IP introduces a 4-byte IP address per network interface per device in addition to the MAC address per interface per device. They can often be redundant. 

Below will show the IPv4 Datagram format. 

In the transmission process, frames are both encapsulated and decapsulated. 

Now, the network layer has 2 responsibilities, Fragmentation/reassembly and routing/forwarding. 

Now, let's move on to IoT. 

The Internet of Things (IoT) is a system of sensors, devices, machines, objects, animals or people that are provided with unique identifiers (UIDs) and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction. Simply put, IoT is about taking all the things in the world and connecting them to the internet. Here is the lifecycle of IoT. 

IoT devices communicate wirelessly, have low power (battery or even no-batteries), low bitrate, and mostly upstream traffic rate. Some advanced IoT devices might have higher power, but it won't be the focus of this article.  

LoRa is the de facto wireless platform of Internet of Things (IoT). Semtech's LoRa chipsets connect sensors to the Cloud and enable real-time communication of data and analytics that can be utilized to enhance efficiency and productivity. The variables are Application payload size of 13 bytes , Header size of 13 bytes, bit rate of 50kbps, Bandwidth of 125kHz, Coding rate 4/5, Preamble symbols 8 and Spreading Factors SF6,…, SF12. Each symbol is spread on air as a longer chip sequence.  The higher the spreading factor, the more the range and the better reception but also more airtime. Here is the Spreading factor with the amount of time per symbol, preamble, and payload. 

The channel gain identifies the amount of signal power degradation (and a phase shift) in a communication channel.

Now let's point out the Channel Gain Computation. Let Px be the transmitted signal power (2.5mW for Bluetooth, 20mW LoRa, , 80mW for WiFi, 100mW for mobile phones, 40W for large base stations). Let Py be the received signal power. Let N0 be the noise power (W/Hz). the channel Gain is ratio of Py to Px. Py/Px = 10 log10(Py/Px) dB. 0 dB means no gain. If Py is greater that means we got positive dB (gain) and vice versa. The power from a transmitting antenna spreads out of the surface of an expanding sphere at the speed of light. As the signal spreads out, the sphere gets larger and the transmitted signal power is further divided on the sphere surface. This surface area is S = 4 or 4πd^2, where d is radius of sphere, the distance between transmitted and receiver antennas. Py depends on the size of the receiving antenna. Size of antenna is A = (λ^2)/4π. We then will get β = 10 log 10 (A/S) dB to determine the channel gain. Wavelength λ given a carrier frequency f is computed as v/f where v is speed of light. Plotted below the channel gain is greater for IoT devices and less for 5G. 

The signal power at the receiver is a much smaller fraction of the transmitted power when using higher frequency carrier waves, so we need to compensate. The fraction also reduces even more rapidly the more distance we put from the Access Point. 

Here is a graph of Range vs Bandwidth. Wi-Fi has a shorter range higher bandwidth. Cellular has a longer range and higher bandwidth. IoT has longer range, lower bandwidth. 

N0 (noise at receiver) is gen measuring the power spectral density at the receiver. A typical value is N- = 10^-14.4 W/Hz and thus overall noise power is N0B = -174 + 10 log10(B) dBm. Thus signal to noise ratio is Py/(N0 * B) or (Px * β)/(N0 * B). Typical B values would be 22MHz for Wifi, 5-24 mHz for 4G, 125-500kHz for LoRa, and 5-100MHz for 5G mmWave. 

Now if I want to plot of graph relating channel capacity and transmit power when β/N0 = 10^6 Hz/W, then we get the following:

In conclusion, increasing the transmit power when SNR is low, leads to pronounced higher capacity. Increasing the power when SNR is already high is not as effective in increasing capacity. The higher the bandwidth, the higher the needed transmit power. Now time to  plot Channel capacity to SNR (Signal to Noise Ratio). 

Increasing the bandwidth when B is scarce leads to pronounced higher capacity. After a threshold, the increase in B is not effective. That is, if transmit power is fixed, an unlimited increase in B is not effective.

The receiver sensitivity, Sy, is the lowest signal power that the receiver can demodulate. The Link margin is defined as M = Py - Sy. For example, if Py = -80 dBm and Sy = 90 dBm then M = 10 dBm = 10 mW. We want to maximize link margin. Signals get weaker in cables, can amplify on antenna.