Micro-magnetic Simulation Analysis Of Shingled Magnetic Recording System

According to the report of International Data Corporation,the global data has exceeded 16 ZB in 2016 and will exceed 160 ZB in 2025.With the explosive growth of digital information,the huge market demand for large-capacity storage has promoted the continuous advancement of traditional magnetic recording and various new types of storage devices.Compared with semiconductors,optical storage,etc.,magnetic storage still has advantages in terms of storage cost,service life and reparability.The hard disks are expected to occupy a major share of the information storage market in the next decades.With the rapid development of magnetic recording technology,the area density of hard disk has increased from 2kbpsi to 1.5Tbpsi~2Tbpsi by Moore's Law in the past.This amazing growth rate is benefited by the use of a series of new technologies.But with the increase of area density,magnetic recording is facing a more and more severe trilemma,so the increase trend is slowing down.In order to break through the bottleneck of area density,the researchers and industry have proposed many new technologies such as exchange coupled composite(ECC)media,bit patterned media,energy assisting magnetic recording,shingled magnetic recording(SMR)and two-dimensional magnetic recording(TDMR).In terms of recording media,the ECC media is composed of a top soft magnetic layer and a bottom hard magnetic layer.The hard layer is used for ensuring thermal stability,and soft layer is used for promoting the switching of hard layer,thereby reducing the switching field of ECC media.In terms of data writing,the SMR which uses a large write head to generate stronger writing field overlaps a portion of adjacent tracks to reduce the track pitch.The SMR does not require a large structural modification of conventional write head to facilitate the industrial upgrading.In terms of data reading,TDMR realizes read-back detection of low SNR data signals under high-density storage from the perspective of signal processing.Through the micro-magnetic simulation,this paper analyzes the writing process of shingled magnetic recording with ECC media,which provides the theoretical support for the optimization of write head design,ECC preparation parameters and TDMR reading process.From the aspects of storage media,write head design and read signal processing,the integrated technology of ECC media,SMR and TDMR improves system performance,providing a theoretical and practical basis for achievingarea density of 10 Tbpsi.Theoretical study and optimization of manufacturing parameters for ECC media.By constructing the three-dimensional Voronoi finite element model of ECC media,the effects of parameters such as anisotropy constant,exchange coupling constant and saturation magnetization on the switching field are analyzed.The preparation parameters of L10-Fe Pt based ECC media are selected and optimized.From the perspective of angular dependence of coercivity,a suitable soft magnetic layer material is selected.The manufacturing size of grain is obtained according to the relationship between switching field and writing ability targeting an area density of 10 Tbpsi.Modeling and optimization of write head for shingled magnetic recording.A method for calculating the erase band width(EBW)of SMR with ECC media is proposed.This method calculates the EBW combined with the write-in error rate and angular dependence of media coercivity.This paper analyzes the influence of skew angle and rounded corner on the write performance.The results show that smaller skew angle could minimize the erase band noise.For different write heads,the appropriate range of skew angle should be set.The lower and upper limits of skew angle is found by taking the tradeoff between adequate writing field and moderate EBW.The optimized rounded write head can increase the field gradient and reduce the fringing field.On the basis of consideration of skew angle and rounder corner effect on write performance,the shingled write head is optimized to achieve ultra-high area density.The read channel of TDMR is modeled,and a logistic regression prediction method for recovering the written data is proposed.The logistic regression prediction model uses the read-back signal and target bit data as the feature vectors and label of sample,respectively.The constrained coding patterns are proposed by statistically analyzing the error conditions in experiments.The results show that the constrained coding patterns can reduce the bit error rate of prediction model.